System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting information from a data source; means for storing the collected information in a database; means for cleaning the stored information and shaping the data; means for analyzing the shaped data using natural language processing techniques to extract important keywords and patterns; means for training a predictive model based on historical data to make future predictions; means for visualizing the prediction results and analysis results in a dashboard format and presenting them to a terminal; and means for analyzing customer feedback and behavioral data and presenting customer satisfaction improvement measures.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In today's business environment, there is a huge amount of diverse information, including customer feedback, behavioral data, and market trends, making it difficult to manually understand this information. Companies also need to make more evidence-based decisions and are being asked to improve efficiency, competitiveness, and customer satisfaction. However, current systems and methods require a lot of time and resources to analyze data, making it difficult to make accurate predictions or respond quickly. There is a need for a system that can solve these issues and enable companies to effectively utilize the vast amount of data. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means. First, it comprises means for collecting information from a data source. Next, it comprises means for storing the collected information in a database, and means for cleaning the stored information and formatting the data. It further comprises means for analyzing the formatted data using natural language processing technology and extracting important keywords and patterns. It comprises means for training a predictive model based on past data and making future predictions, and means for visualizing the prediction results and analysis results in dashboard format and presenting them on a terminal. Finally, it comprises means for analyzing customer feedback and behavioral data and proposing measures to improve customer satisfaction. In this way, it provides a system that eliminates the difficulty of understanding information due to the vast amount of data and supports future predictions and decision-making, thereby improving corporate efficiency and competitiveness and contributing to increased customer satisfaction.

[0006] 1. "Data Source" refers to an external or internal source from which information is collected, such as social media, customer review sites, or an internal CRM system.

[0007] 2. "Means" refers to the method or technique used to achieve a specific purpose, including specific devices or programs required to perform each step of the invention.

[0008] 3. A "database" is a systematic collection of organized information that is designed to be easily searchable and manageable.

[0009] 4. "Cleaning" refers to the process of removing or correcting inaccurate, incomplete, or unnecessary data to improve data quality.

[0010] 5. "Data shaping" refers to the process of organizing and adjusting the format and structure of data to make it easier to analyze.

[0011] 6. "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language, including text analysis and sentiment analysis.

[0012] 7. A "token" is a basic unit in natural language processing: a word, phrase, or symbol that is broken down when parsing a sentence.

[0013] 8. "Predictive model" refers to an algorithm or mathematical model used to predict future events based on historical data.

[0014] 9. "Dashboard" means an interface that provides a visual representation of data and allows users to quickly and easily grasp important information.

[0015] 10. "Customer Satisfaction Improvement Plan" means a specific action plan to improve products and services based on customer needs and preferences and increase customer satisfaction. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The present invention is a system that enables companies to efficiently process huge amounts of data and support future predictions and decision-making. The program processing of this system will be specifically described below.

[0038] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[0039] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[0040] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[0041] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[0042] The server visualizes the prediction and analysis results in a dashboard format and displays them on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[0043] Furthermore, the server analyzes customer feedback and behavioral data to suggest measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user, allowing the user to take appropriate actions to increase customer satisfaction.

[0044] Specific examples

[0045] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[0046] The server then analyzes customer feedback using NLP techniques, performing sentiment analysis and topic modeling, analyzing behavioral data and conducting clustering to identify customer purchasing patterns, which allows the server to identify desired features and shortcomings of new products.

[0047] The server trains a predictive model based on past data to forecast demand for new products. For example, it can estimate sales for the next three months. The forecast and analysis results are visualized in a dashboard format and displayed on the device.

[0048] Users can check the results through a dashboard and develop specific marketing strategies and production plans, taking into account the insights provided by the server. Furthermore, the server analyzes customer feedback and provides specific suggestions for improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[0049] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information due to the vast amount of data, make evidence-based decisions, and improve efficiency and competitiveness. It also contributes to improving customer satisfaction.

[0050] The processing flow will be explained below.

[0051] Step 1:

[0052] The server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server uses a data collection API to retrieve data from each source and stores it in a database.

[0053] Step 2:

[0054] The server cleans the data. The collected data may contain missing values ​​or outliers. The server detects these data and either fills in the missing values ​​with the average value or corrects the outliers. The server checks the data consistency to improve the data quality.

[0055] Step 3:

[0056] The server formats the data. To format the data in a form that is easy to analyze, it selects features and normalizes the data. In this step, the server extracts the necessary columns and standardizes the scale of the numerical data.

[0057] Step 4:

[0058] The server uses natural language processing technology to analyze the text data. For example, it tokenizes customer feedback and extracts important keywords. It also performs sentiment analysis and topic modeling to understand trends in customer opinions and emotions.

[0059] Step 5:

[0060] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify user segments and purchasing patterns. It uses clustering algorithms to identify groups of users with similar behavior.

[0061] Step 6:

[0062] The server trains the predictive model. It trains time series analysis models and regression analysis models based on past data to predict future market demand and customer behavior. In this step, data is trained and tested to evaluate the performance of the model.

[0063] Step 7:

[0064] The server performs the prediction execution, using the trained model to predict future events, such as estimating sales figures or customer intent after a new product launch.

[0065] Step 8:

[0066] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts to create an interactive dashboard.

[0067] Step 9:

[0068] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[0069] Step 10:

[0070] The server presents measures to improve customer satisfaction. It analyzes customer feedback and behavioral data and makes specific suggestions for improving products and services. Users can take appropriate actions based on the suggestions presented via their devices.

[0071] Example 1

[0072] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0073] Modern companies need to collect huge amounts of data from social media, customer review sites, and other sources, but it is difficult to efficiently organize and analyze that data and use it for future predictions and decision-making. Furthermore, there is a lack of technology to automatically generate specific proposals to improve customer satisfaction by utilizing customer feedback and behavioral data, so these issues must be resolved.

[0074] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0075] In this invention, the server includes: means for collecting information from data sources such as social media, customer review sites, and internal CRM systems; means for converting the collected information into JSON or CSV format and storing it in a database; means for cleaning the stored information using a library such as Pandas and formatting the data; means for analyzing the formatted data using natural language processing technology to extract important keywords, sentiment, and topics; means for the server to analyze behavioral data using techniques such as K-means clustering and identify user segments and purchasing patterns; means for training ARIMA models and regression analysis using past data to predict future market demand and customer behavior; means for visualizing the prediction and analysis results in dashboard format using Matplotlib or Plotly and displaying them on a terminal; and means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction. This enables efficient processing of massive amounts of data, supporting corporate decision-making, and improving customer satisfaction.

[0076] A "data source" is the original location or medium where information is stored, such as social media, customer review sites, or internal CRM systems.

[0077] "Gathering" refers to the process by which the server obtains the required information from the data sources.

[0078] "JSON" and "CSV" are standard data formats for storing collected data.

[0079] A "database" is a storage system that systematically organizes and stores information, allowing it to be accessed and managed efficiently.

[0080] "Cleaning" is the process of correcting or removing missing or outlier values ​​in data.

[0081] The "Pandas library" is a library for data processing and analysis in the Python language.

[0082] "Formatting" is the process of converting data into a form that is easier to analyze.

[0083] "Natural language processing technology" refers to the technology that enables computers to understand and process human language.

[0084] "Keywords" are important words and phrases extracted from the data.

[0085] "Sentiment analysis" is the process of identifying and classifying emotions and opinions contained within text data.

[0086] "Topic modeling" is a technology that automatically extracts topics from large amounts of text data.

[0087] "K-means clustering" is a machine learning algorithm for dividing data into several clusters (groups).

[0088] A "user segment" is a collection or group of users who share common characteristics.

[0089] "Purchasing patterns" are behaviors and tendencies that can be identified from a user's past purchasing history.

[0090] The "ARIMA model" is a statistical model used to analyze and predict time series data.

[0091] "Regression analysis" is a statistical method for filling in the relationships between variables.

[0092] "Forecasting" is the estimation of future events based on past data.

[0093] "Visualization" refers to the visual display of data and its analysis results using graphs and charts.

[0094] A "dashboard" is an interface for displaying multiple pieces of information on a single screen.

[0095] "Customer feedback" refers to opinions and evaluations from customers.

[0096] "Measures to improve customer satisfaction" are specific measures and proposals to increase customer satisfaction.

[0097] This invention is a system that enables companies to efficiently process large amounts of data and support future predictions and decision-making. This system is composed of a server, terminals, and users.

[0098] Server Processing

[0099] The server first collects information from data sources such as social media, customer review sites, and internal CRM systems. This information is collected using scripts and APIs written in Python. For example, to use the Twitter API to collect tweets containing a specific hashtag, the following method is used:

[0100] The server converts the collected data into JSON or CSV format and stores it in a database. This unifies the information collected from different data sources and makes subsequent processing easier. The server also uses the Pandas library to clean the data, correcting missing values ​​and outliers.

[0101] Data Shaping and Analysis

[0102] The server then formats the data and converts it into a format suitable for analysis. For example, it tokenizes the text data of customer reviews and removes stop words. Using this formatted data, the server analyzes customer feedback using natural language processing techniques. This processing involves keyword extraction using Scikit-learn's TfidfVectorizer and sentiment analysis using NLTK.

[0103] Clustering and Prediction

[0104] The server then uses algorithms such as K-means clustering to analyze behavioral data and identify user segments and purchasing patterns. It also uses historical data to train ARIMA models and regression analysis to predict future market demand and customer behavior. For example, an ARIMA model is used to forecast sales for the next three months based on past sales data.

[0105] Visualization and Presentation

[0106] The server visualizes the prediction and analysis results in dashboard format using Matplotlib or Plotly. This allows the data sent to the terminal to be displayed in a visually easy-to-understand format. Users can use this dashboard to check the analysis and prediction results and develop specific marketing strategies and production plans. The server also reanalyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction to users. This allows users to easily receive suggestions for product improvements and new services.

[0107] Specific examples

[0108] For example, consider a company planning to launch a new product. The server uses the Twitter API to collect tweets containing the hashtag "new product" and stores them in a database. The Pandas library is then used to impute missing values ​​in the collected data, tokenize it, and remove stop words. Next, Scikit-learn and NLTK are used to extract important keywords and perform sentiment analysis. K-means is used for clustering, and an ARIMA model is used to predict sales for the next three months. These results are visualized with Matplotlib and sent to the device.

[0109] Prompt Sentence Examples

[0110] "Write a program that uses the Twitter API to collect tweets containing a specific hashtag, cleans the data using Pandas, extracts important keywords using TfidfVectorizer, and then uses an ARIMA model to forecast sales for the next three months."

[0111] As a result, this system can efficiently process huge amounts of data, support corporate decision-making, and improve customer satisfaction.

[0112] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0113] Step 1:

[0114] The server collects data from social media, customer review sites, and internal CRM systems.

[0115] Input: Twitter API, customer review site API, data access method for internal CRM system

[0116] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "new product." Next, it collects product reviews from customer review sites via the API and retrieves customer purchase history data from the internal CRM system.

[0117] Output: Collected raw data (JSON and CSV formats)

[0118] Step 2:

[0119] The server stores the collected data in a database.

[0120] Input: Collected raw data (JSON or CSV format)

[0121] Specific operation: The server converts the collected data into JSON or CSV format and saves it in the corresponding table in the database. For example, Twitter data is saved in a table called "SNS_data".

[0122] Output: Data stored in a database in a uniform format

[0123] Step 3:

[0124] The server cleans the data.

[0125] Input: Data stored in a database

[0126] Specific operation: The server uses the Pandas library to detect, correct, or remove missing and outlier values. For example, in a customer review dataset, it imputes missing values ​​with the mean value and filters out outliers.

[0127] Output: Cleaned dataframe

[0128] Step 4:

[0129] The server formats the data and converts it into a format suitable for analysis.

[0130] Input: Cleaned dataframe

[0131] What it does: The server tokenizes the customer review text data and removes stop words. Specifically, it vectorizes the data using Scikit-learn's TfidfVectorizer.

[0132] Output: Tokenized text data with stop words removed

[0133] Step 5:

[0134] The server analyzes the data using natural language processing technology.

[0135] Input: Tokenized text data with stop words removed

[0136] How it works: The server extracts important keywords, performs sentiment analysis using NLTK, and extracts major topics using LDA topic modeling.

[0137] Output: Keyword list, sentiment score, topic list

[0138] Step 6:

[0139] The server clusters the behavioral data.

[0140] Input: Organized behavioral data

[0141] How it works: The server uses the K-means clustering algorithm to group users' purchasing patterns, for example, dividing users into multiple clusters based on purchase frequency or purchase amount.

[0142] Output: User's cluster information

[0143] Step 7:

[0144] The server uses historical data to train the predictive model.

[0145] Input: Existing sales data and customer behavior data

[0146] What it does: The server uses an ARIMA model to forecast sales for the next three months, and also trains the model to forecast market demand using regression analysis.

[0147] Output: Sales forecast results, demand forecast results

[0148] Step 8:

[0149] The server visualizes the prediction and analysis results.

[0150] Input: Sales forecast results, keyword list, sentiment score, user cluster information

[0151] Specific operation: The server uses Matplotlib or Plotly to visualize the prediction results as line graphs, pie charts, and bar graphs.

[0152] Output: Visualized graphs and charts

[0153] Step 9:

[0154] The terminal presents the visualized results to the user.

[0155] Input: Visualized graphs and charts

[0156] Specific operation: The terminal launches the dashboard and displays visualized data on the screen. Users can access the dashboard via a PC or tablet and check the analysis results.

[0157] Output: Analysis results displayed in a dashboard

[0158] Step 10:

[0159] The server then re-analyzes customer feedback and behavioral data to generate suggestions for improvements and new services.

[0160] Input: Customer feedback, behavioral data

[0161] How it works: The server analyzes negative customer feedback and uses machine learning algorithms to identify areas for product improvement. It also automatically generates new service proposals.

[0162] Output: List of improvement proposals, new service proposals

[0163] The above are the specific processing steps of this system.

[0164] (Application example 1)

[0165] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0166] In recent years, companies have been able to obtain vast amounts of data, but there is a demand for efficient systems that can effectively utilize this data and help with business decision-making. In particular, brick-and-mortar stores lack the means to quickly analyze data collected in real time and develop sales forecasts and marketing strategies. Furthermore, there is a demand for rapid response by providing store staff with analysis of customer feedback and demand forecasts in real time. To address these challenges, the present invention provides advanced data processing technology and a means for visualizing data in real time.

[0167] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0168] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and shaping the data, means for analyzing the shaped data using natural language processing technology and extracting important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction, means for cleaning the collected data in real time and performing analysis, sentiment analysis, and topic modeling using natural language processing technology, means for making demand forecasts based on past sales data and providing sales forecasts for the next period, and means for displaying data visualizations on smart glasses and notifying important trends and alerts in real time. This enables companies to extract useful information from massive amounts of data and use it for customer analysis, demand forecasting, and sales management.

[0169] A "data source" is the source of the original data used to collect information, and includes social media sites and customer review sites.

[0170] "Means for collecting information" refers to the techniques and methods for efficiently acquiring the necessary information from data sources.

[0171] A "database" is a collection of structured data that is used to systematically store and manage collected information.

[0172] "Information cleaning methods" are techniques and methods for correcting missing data and outliers from collected data and improving the quality of the data.

[0173] "Means for data preparation" refers to techniques and methods for converting cleaned data into an analyzable form and formatting it appropriately.

[0174] "Natural language processing technology" is a computer technology for processing human language, including text tokenization, sentiment analysis, and topic modeling.

[0175] "Means for extracting important keywords and patterns" refers to techniques and methods for finding important information and trends that are useful for business from analyzed data.

[0176] A "means for training a predictive model" is a technique or method for building and training a model to predict future events or conditions based on past data.

[0177] "Means for making future predictions" are technologies or methods that use trained predictive models to forecast future demand, sales, etc.

[0178] "Means for visualizing in dashboard format" refers to techniques and methods for displaying analysis results and prediction results in a format that is intuitively easy to understand, including graphs and charts.

[0179] "Means for displaying on a device" refers to the technology or method for displaying the information visualized in dashboard format on a device used by the user (e.g., smart glasses, tablet).

[0180] "Means for analyzing customer feedback and behavioral data" refers to techniques and methods for analyzing customer opinions and behavioral history to evaluate customer satisfaction and behavioral trends.

[0181] "Means for proposing measures to improve customer satisfaction" refers to techniques and methods that propose specific actions to improve customer satisfaction based on customer feedback and behavioral data.

[0182] "Means for cleaning data collected in real time" refers to techniques and methods for immediately correcting missing or outliers in data collected in real time, thereby maintaining the quality of the data.

[0183] "Means for sentiment analysis and topic modeling" refers to techniques and methods that use natural language processing techniques to analyze sentiment trends and major topics from text.

[0184] "Means for forecasting demand based on past sales data" refers to techniques and methods for forecasting future sales and demand using historical sales data.

[0185] The "means for providing a sales forecast for the next period" refers to a technique or method for predicting future sales for a specific period and providing the results.

[0186] "Means for displaying on smart glasses" means the technology or method for displaying visualized data or alerts on smart glasses, allowing users to view the information in real time.

[0187] "Means for notifying important trends and alerts" refers to technologies and methods that immediately notify users of trends or anomalies that require particular attention based on the results of data analysis.

[0188] The "data analysis system for brick-and-mortar stores" of the present invention effectively collects huge amounts of data and analyzes it in real time to optimize brick-and-mortar store operations. Specific embodiments of this system are described below.

[0189] 1. Data Collection

[0190] The server collects customer feedback and sales data from social media, customer review sites, internal CRM systems, etc. For example, it uses the Twitter API to retrieve tweets related to product names and collect customer opinions.

[0191] 2. Data cleaning and formatting

[0192] The collected data is cleaned using the Pandas library to correct missing data and outliers and improve data quality. This cleaned data is then formatted using natural language processing techniques, specifically NLTK and SpaCy, to tokenize the text and convert it into a parseable format.

[0193] 3. Data analysis

[0194] The server uses natural language processing technology to perform sentiment analysis and topic modeling on the formatted data. For example, it analyzes customer tweets and determines whether their content is positive or negative. This allows for a quantitative understanding of product reviews on social media.

[0195] 4. Demand forecasting

[0196] Based on past sales data, the server uses Scikit-learn to train a predictive model. For example, monthly sales data is used to forecast sales for the next three months. The trained model can then predict future demand with high accuracy.

[0197] 5. Data Visualization and Notification

[0198] The server visualizes the prediction and analysis results in the form of a dashboard. Graphs and charts are created using Matplotlib and Seaborn. The generated visualization data is displayed in real time on smart glasses via AR Kit or the Google® Glass® API. This allows users to receive the information they need to run their stores in real time and instantly grasp important trends and alerts.

[0199] Specific examples

[0200] As a concrete example, consider a store planning to launch a new product. In this system, the server collects information about the new product from social media and customer review sites and performs sentiment analysis. For example, if a prompt such as "Collect social media reactions about product X and display the results of the sentiment analysis on the smart glasses in real time" is input into the generative AI model, the results of the sentiment analysis will be displayed in real time on the smart glasses. In addition, sales forecasts for the next three months are generated based on past sales data and notified to the user. This can be used to develop marketing strategies and production plans for the new product.

[0201] The present invention is a system that eliminates the difficulty of understanding information due to the vast amount of data and provides the ability to make evidence-based decisions, thereby significantly improving the competitiveness of physical stores.

[0202] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0203] Processing Steps

[0204] Step 1:

[0205] The server collects information from data sources (social media, customer review sites, internal CRM systems, etc.).

[0206] Input: API key or authentication information for data collection, search keywords.

[0207] Output: The raw data collected.

[0208] Specific operation: The server uses the Twitter API to search for tweets related to "product name" and obtains the results in text format.

[0209] Step 2:

[0210] The server stores the collected raw data in a database.

[0211] Input: Collected raw data (text data).

[0212] Output: The saved database entries.

[0213] Specific operation: The server converts the tweet data into a tabular format using a Pandas data frame and saves it in an SQL database.

[0214] Step 3:

[0215] The server cleans the stored data and corrects missing or outlier values.

[0216] Input: A saved database entry.

[0217] Output: Cleaned data.

[0218] Specific operation: The server uses the Pandas library to remove or impute missing data and correct outliers.

[0219] Step 4:

[0220] The server formats the cleaned data and applies natural language processing techniques.

[0221] Input: Cleaned data.

[0222] Output: Formatted text data.

[0223] Specific operation: The server uses the NLTK or SpaCy library to tokenize the text data, remove stop words, and extract the necessary features.

[0224] Step 5:

[0225] The server analyzes the data using natural language processing techniques to perform sentiment analysis and topic modeling.

[0226] Input: Formatted text data.

[0227] Output: Analysis results (sentiment scores, topic distribution).

[0228] Specific operation: The server performs sentiment analysis of the text using NLTK's SentimentIntensityAnalyzer and extracts important topics using an LDA model.

[0229] Step 6:

[0230] The server trains a predictive model based on past sales data and makes future predictions.

[0231] Input: Historical sales data, cleaned data.

[0232] Output: Sales forecast results.

[0233] What happens: The server uses Scikit-learn to train a linear regression model and perform sales forecasts for the next three months.

[0234] Step 7:

[0235] The server visualizes the prediction and analysis results in dashboard format and displays them on the terminal.

[0236] Input: Sales forecast results, sentiment analysis and topic modeling results.

[0237] Output: Visualized graphs on a dashboard.

[0238] Specific operation: The server creates graphs and charts using Matplotlib and Seaborn and generates a web dashboard.

[0239] Step 8:

[0240] The server analyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction.

[0241] Input: Customer feedback data, behavioral data.

[0242] Output: Specific action proposals to improve customer satisfaction.

[0243] Specific operation: The server performs sentiment analysis based on the feedback data, extracts areas for improvement in products and services, and notifies the marketing team.

[0244] Step 9:

[0245] The server cleans the collected data in real time and uses natural language processing technology to perform analysis, sentiment analysis and topic modeling.

[0246] Input: Raw data collected in real time.

[0247] Output: Cleaned and analyzed data.

[0248] Specific operation: The server cleans real-time streaming data with Pandas, analyzes sentiment scores with NLTK, and extracts topics with an LDA model.

[0249] Step 10:

[0250] The server makes a demand forecast based on past sales data and provides a sales forecast for the next period.

[0251] Input: Historical sales data.

[0252] Output: Sales forecast for next period.

[0253] What it does: The server uses Scikit-learn's time series forecasting model to forecast future sales and generate results.

[0254] Step 11:

[0255] The server displays data visualizations on the smart glasses, notifying users of important trends and alerts in real time.

[0256] Inputs: Visualization data, important trends or alerts.

[0257] Output: Real-time data displayed on smart glasses.

[0258] What it does: The server uses AR Kit or Google Glass APIs to display important data on the smart glasses and notify you of trends and alerts.

[0259] Through the above steps, this system can effectively collect and analyze huge amounts of data, significantly supporting decision-making in brick-and-mortar store operations.

[0260] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0261] The present invention is a system that aims to enable companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, making it possible to present more sophisticated measures to improve customer satisfaction. The program processing of this system is explained in detail below.

[0262] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[0263] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[0264] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[0265] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[0266] The server visualizes the prediction and analysis results in the form of a dashboard and displays it on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[0267] Furthermore, in this invention, the server analyzes customer feedback and behavioral data to present measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user. Here, the emotion engine plays a particularly important role. The emotion engine recognizes the user's emotions and stores the emotion data in a database. The server also analyzes this emotion data to assist in providing more precise measures to improve customer satisfaction.

[0268] Specific examples

[0269] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[0270] The server uses NLP techniques to analyze customer feedback, perform sentiment analysis and topic modeling, analyze behavioral data, and perform clustering to identify customer buying patterns, which allows the server to identify desired features and shortcomings of new products.

[0271] The server then trains a predictive model, forecasting demand for new products based on past data and estimating sales figures for the next three months, for example. The forecast and analysis results are then visualized in a dashboard format and displayed on the device.

[0272] Users can check the results through the dashboard and develop specific marketing strategies and production plans, while also referring to the data from the emotion engine provided by the server. The emotion engine analyzes customer emotions, and the server uses this data to suggest specific improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[0273] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information caused by vast amounts of data, make evidence-based decisions, and improve efficiency and competitiveness.In addition, by utilizing the emotion engine, it can further contribute to improving customer satisfaction.

[0274] The processing flow will be explained below.

[0275] Step 1:

[0276] The server collects information from data sources, such as social media, customer review sites, and internal CRM systems. The server uses data collection APIs to retrieve data from each source and stores it in a database.

[0277] Step 2:

[0278] The server cleans the stored data, removing missing data and noise, and detecting and correcting outliers, for example by imputing missing values ​​with the mean or correcting outliers to improve the quality of the data.

[0279] Step 3:

[0280] The server formats the data, converts it into a format that is easy to analyze, extracts the necessary features, and normalizes the data so that all data is handled on a uniform scale.

[0281] Step 4:

[0282] The server uses natural language processing (NLP) to analyze the text data, tokenize customer feedback, extract key keywords and phrases, and perform sentiment analysis and topic modeling to understand trends in customer opinions and sentiment.

[0283] Step 5:

[0284] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify patterns in user behavior. It uses a clustering algorithm to group users with similar behavior.

[0285] Step 6:

[0286] The server uses an emotion engine to recognize the user's emotions. For example, it analyzes emotions from customer text feedback and facial expression data. The emotion data is stored in a database and integrated with other data for analysis.

[0287] Step 7:

[0288] The server trains the predictive model. Based on past data, it uses time series analysis and regression analysis models to predict future market demand and customer behavior. The predictive model is then evaluated for accuracy through training and testing on data.

[0289] Step 8:

[0290] The server performs the prediction execution, using the trained model to make specific predictions about future events, such as estimating sales figures or customer intent after a new product launch.

[0291] Step 9:

[0292] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts and displays the information on an interactive dashboard.

[0293] Step 10:

[0294] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[0295] Step 11:

[0296] The server proposes measures to improve customer satisfaction based on data from the emotion engine. It comprehensively analyzes customer feedback, behavioral data, and recognized emotional data to propose specific improvements to products and services. The proposals are presented via the device, and the user can then take action to increase customer satisfaction.

[0297] Example 2

[0298] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0299] Modern companies face challenges in efficiently collecting massive amounts of data, analyzing it, and making future predictions and decisions. Furthermore, a lack of precise ways to recognize customer sentiment and improve customer satisfaction makes it difficult to properly understand customer needs and respond quickly. Existing systems perform data cleaning, natural language processing analysis, and predictive model training separately, making integrated analysis and prediction difficult, and making effective decision-making difficult.

[0300] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0301] In this invention, the server includes: [means for collecting information from data sources;] [means for storing the collected information in a database;] [means for cleaning the stored information and formatting the data;] [means for analyzing the formatted data using natural language processing technology and extracting important keywords and patterns;] [means for training a predictive model based on past data and making future predictions;] [means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal; and [means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction using an emotion engine.] This enables companies to collect and analyze data in an integrated manner, make future predictions and decisions, and quickly and effectively respond to customer needs and improve customer satisfaction.

[0302] "Data sources" are sources of data collected from external and internal systems and platforms, such as social media, customer review sites, and CRM systems.

[0303] A "database" is an information system that stores collected information and enables it to be efficiently managed and searched. Examples include relational databases and NoSQL databases.

[0304] "Cleaning" is the process of correcting and removing missing data and outliers from stored information to improve data quality.

[0305] "Natural language processing technology" refers to computer techniques for analyzing, understanding, and generating human language. Specifically, it includes tokenization, sentiment analysis, and topic modeling.

[0306] A "predictive model" is a statistical or machine learning model trained to predict future trends based on past data. It includes techniques such as time series analysis and regression analysis.

[0307] A "dashboard" is an interface that visually displays analysis and prediction results, allowing users to intuitively grasp the situation.

[0308] An "emotion engine" is a technology or algorithm that analyzes user emotions and generates corresponding emotional data. It helps to precisely improve customer satisfaction.

[0309] "Customer satisfaction improvement measures" are specific action plans for presenting improvements and new proposals for products and services based on customer evaluations and feedback.

[0310] The present invention is a system that enables companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, by incorporating an emotion engine that recognizes user emotions, it is possible to provide more sophisticated measures to improve customer satisfaction. Specific embodiments of this system are described below.

[0311] First, the server collects information from data sources. Data sources include social media, customer review sites, and internal CRM systems. For example, the server might use an API to collect data related to a specific hashtag from a social media site. The server then converts the collected information into an appropriate format and stores it in a database. Relational databases or NoSQL databases are often used for this purpose.

[0312] As a concrete example, a server uses the Twitter API to collect tweets containing the hashtag "new product release," converts them into JSON format, and stores them in a database.

[0313] Next, the server cleans the stored information. This step uses Python's Pandas library to impute missing data and correct outliers. For example, records with missing purchase dates are imputed with the user's closest purchase date. Additionally, any outliers that exist are removed through manual review or automated correction.

[0314] The server then formats the data, selects and normalizes the necessary features, for example, converting customer age, purchase frequency, and purchase amount to a standardized scale using scaling and feature selection techniques. The formatted data is then split into a training dataset and a test dataset.

[0315] The server then uses natural language processing (NLP) techniques to analyze this data. Specifically, it uses the SpaCy library to tokenize customer feedback and extract key keywords. It also performs sentiment analysis to assign a positive, negative, or neutral sentiment score to the feedback. Finally, it uses Scikit-learn to cluster behavioral data and identify customer purchasing patterns.

[0316] Next, the server uses TENSORFLOW (registered trademark) to train a predictive model for time series analysis. For example, it predicts sales figures for the next three months based on past data. Evaluation indicators are set for the predictive model, and reliability is evaluated. If an abnormal value is detected, the model can be retrained.

[0317] The server uses Tableau to visualize these predictions and analysis results in dashboard format and displays them on the device. Users can access the dashboard using a PC or tablet and intuitively grasp the situation, which is useful for formulating marketing strategies and production plans.

[0318] Furthermore, the server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The emotion engine analyzes the user's emotions and stores the results in a database. The server then generates measures to improve customer satisfaction based on the emotion data and proposes them to the user.

[0319] Examples of prompt statements

[0320] How do I impute missing data and remove outliers using Python's Pandas?

[0321] "Describe an effective method for conducting sentiment analysis from customer feedback using natural language processing."

[0322] "How do I train a time series analysis model using TensorFlow to forecast sales for the next three months?"

[0323] "Please explain how to visualize data collected with Tableau in the form of a dashboard."

[0324] This allows companies to overcome the difficulty of comprehending information caused by the sheer volume of data, make evidence-based decisions, and quickly implement concrete actions to improve customer satisfaction.

[0325] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0326] Step 1:

[0327] The server collects information from data sources, including social media, customer review sites, and internal CRM systems. It retrieves data as input by sending API requests. Specifically, it uses the Twitter API to collect tweets containing the hashtag "new product release" and converts the retrieved data into JSON format. It generates a JSON object of the collected data as output.

[0328] Step 2:

[0329] The server stores the collected information in a database. The input is the JSON object collected in step 1. Specifically, it executes an INSERT query against an SQL or NoSQL database to store the data. The output is the number of records stored in the database.

[0330] Step 3:

[0331] The server cleans the stored information. The input is raw data read from the database. Specifically, it uses Python's Pandas library to impute missing data and correct outliers. For example, it imputes missing purchase dates with the user's closest purchase date. The output is a cleaned data frame.

[0332] Step 4:

[0333] The server formats the cleaned data. The input is a cleaned data frame. Specifically, features such as customer age, purchase frequency, and purchase amount are selected and normalized. Feature scaling is performed using Scikit-learn's StandardScaler. The output is a formatted data frame.

[0334] Step 5:

[0335] The server uses natural language processing techniques to analyze the formatted data. The input is a formatted data frame. Specifically, the SpaCy library is used to tokenize the customer feedback and extract important keywords. A sentiment analysis engine is used to calculate a sentiment score for the feedback. The output is a new data frame containing the analysis results.

[0336] Step 6:

[0337] The server clusters the behavioral data. The input is a data frame analyzed using natural language processing techniques. Specifically, it uses Scikit-learn to perform K-means clustering to identify customer purchasing patterns. The output is the data points corresponding to each cluster and the coordinates of the cluster center.

[0338] Step 7:

[0339] The server trains a predictive model. The input is clustered historical data. TensorFlow is used to train a predictive model for time series analysis. For example, a model is created to predict sales for the next three months. The output is a dataset containing the predicted values.

[0340] Step 8:

[0341] The server visualizes the prediction and analysis results in the form of a dashboard and presents it to the terminal. The input is a dataset containing the prediction values ​​and analysis results. Specifically, the dashboard is created using Tableau and provided to the end user through a web-based interface. The output is a visualized dashboard that can be viewed on the terminal.

[0342] Step 9:

[0343] The server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The input is customer feedback data. Specifically, the emotion engine analyzes user feedback and generates an emotion score. This is then stored in a database, and measures to improve customer satisfaction are generated based on the analysis results. The output is a report proposing measures to improve customer satisfaction.

[0344] Through these steps, the system enables businesses to efficiently collect and analyze data and provide actionable insights, which in turn contributes to improving customer satisfaction.

[0345] (Application example 2)

[0346] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0347] In recent years, companies have been required to quickly and accurately collect and analyze massive amounts of data and make appropriate decisions based on that data. However, due to the sheer volume and diversity of data, there is a lack of concrete methods for how to collect, analyze, and utilize this information. Furthermore, while there is a need to further improve customer satisfaction by recognizing customer emotions in real time, current systems have difficulty doing this efficiently. This creates the challenge of being unable to grasp the detailed needs and emotions of customers, making it difficult to present appropriate product proposals and improvement measures.

[0348] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0349] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and formatting the data, means for analyzing the formatted data using natural language processing technology to extract important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, means for analyzing customer reviews and emotion data in real time and recommending products based on related topics and emotion scores, and means for visualizing emotion score distributions and clustering results on the dashboard. This enables companies to efficiently collect and analyze massive amounts of data, recognize customer emotions in real time, and propose product suggestions and improvement measures.

[0350] A "data source" is a data provider that includes a file, database, or other source of information.

[0351] "Means for collecting information" refers to the tools and techniques used to obtain the necessary data from designated data sources and input it into the system.

[0352] "Means for storing information in a database" refers to the technology or system for storing collected data in an appropriate format as structured data.

[0353] "Information cleaning and data formatting methods" are processes and techniques that correct incomplete or erroneous data and convert it into a coherent form.

[0354] "Natural language processing technology" is a technology that allows computers to understand, process, and analyze human language.

[0355] A "predictive model" is a mathematical or statistical model used to predict future behavior or outcomes based on past data.

[0356] A "dashboard format" is an interface or layout that displays data in a visually easy-to-understand way.

[0357] "Customer feedback" refers to customer evaluations and opinions about products and services.

[0358] "Customer behavioral data" refers to data related to customer behavior, such as purchasing history and website browsing history.

[0359] "Customer reviews" are ratings and comments written by customers about products or services.

[0360] "Emotional data" is emotional information extracted from customer reviews and feedback.

[0361] "Real-time analysis" is the process of analyzing data immediately at the moment it is generated.

[0362] "Relevant topics" are the main discussion themes or topics that emerge from the analyzed data.

[0363] An "emotion score" is a numerical representation of the degree of positivity or negativity of emotions extracted from text data.

[0364] "Clustering" is a technique for grouping data based on similarity.

[0365] The present invention is a system for businesses to efficiently collect and analyze data and improve customer satisfaction. The system incorporates data collection, cleaning, natural language processing, predictive model training, visualization with dashboards, sentiment analysis, and clustering functions.

[0366] System Configuration

[0367] 1. Data Collection and Storage

[0368] The server collects data from multiple data sources (e.g., social media, customer review sites, CRM systems, etc.) and stores it in a database. This data collection is done using the requests library.

[0369] 2. Data cleaning and formatting

[0370] The server cleans and formats the collected data. This step involves correcting missing data and outliers to improve data quality. This is done using the Python standard library and Pandas.

[0371] 3. Natural Language Processing and Sentiment Analysis

[0372] The server analyzes the cleaned data using natural language processing (NLP) techniques. Specifically, it performs sentiment analysis using TextBlob to extract important keywords and patterns. It also uses KMeans from scikit-learn for clustering.

[0373] 4. Training a predictive model and forecasting the future

[0374] The server uses historical data to train predictive models that forecast future market demand and customer behavior, using algorithms such as time series analysis and regression analysis.

[0375] 5. Visualizing and Presenting Results

[0376] The prediction and analysis results are visualized in a dashboard format. Users can access the dashboard and check the results using a PC or tablet. Visualization is performed using libraries such as Matplotlib and Dash.

[0377] 6. Proposing measures to improve customer satisfaction

[0378] The server analyzes customer feedback and behavioral data to suggest satisfaction improvement measures tailored to each individual customer. The emotion engine analyzes customer reviews and emotion data in real time and recommends products based on related topics and emotion scores.

[0379] Specific examples

[0380] For example, when a user of the "Online Shopping Improvement Assistant" on the online shopping site views a specific product, such as a "Bluetooth speaker," the server analyzes the reviews on that page. If there are many negative sentiments based on sentiment analysis, the server automatically notifies the user about areas that need improvement. Related topics (e.g., sound quality, battery life, connectivity) extracted from the reviews through clustering are also presented.

[0381] Prompt Sentence Examples

[0382] "Write a Python program that extracts negative sentiment from Bluetooth speaker reviews and notifies users with related topics. Use the TextBlob and scikit-learn libraries."

[0383] These systems enable companies to effectively utilize vast amounts of data and provide specific product suggestions and improvement measures based on customer sentiment, while users can check the analysis results in real time and make decisions quickly.

[0384] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0385] Program processing steps

[0386] Step 1:

[0387] The server collects information from data sources. Specifically, it collects data from multiple data sources (such as social media, customer review sites, and internal CRM systems) through APIs. It primarily uses the requests library to retrieve data in JSON format. The input is the URL and API key of the data source, and the output is the collected raw data.

[0388] Step 2:

[0389] Store the collected information in a database. The server converts the raw data into the required format and stores it in an appropriate database (e.g., MySQL or MongoDB). The input is the collected raw data, and the output is structured data stored in the database.

[0390] Step 3:

[0391] The stored information is cleaned and the data is formatted. The server uses the Pandas library to correct incomplete data and outliers and standardize the data format. The input is raw data retrieved from the database, and the output is cleaned, high-quality data.

[0392] Step 4:

[0393] The formatted data is analyzed using natural language processing techniques. The server uses the TextBlob library to analyze the text data of each review and calculate a sentiment score. It also performs clustering using scikit-learn. The input is the cleaned data, and the output is the sentiment score and clustering results for each review.

[0394] Step 5:

[0395] The server trains a predictive model based on historical data to predict market demand and customer behavior using a predictive algorithm. The input is the training data, and the output is the trained predictive model.

[0396] Step 6:

[0397] The prediction and analysis results are visualized in dashboard format and displayed on the terminal. The server uses libraries such as Matplotlib and Dash to visualize the sentiment scores, clustering results, and prediction results in graphs and charts. The input is the prediction model and analysis data, and the output is a visually easy-to-understand dashboard.

[0398] Step 7:

[0399] Customer feedback and behavioral data are analyzed to suggest measures to improve customer satisfaction. The server generates a specific action plan to improve customer satisfaction based on the analysis results and presents it to the user. The input is feedback data and behavioral data, and the output is a detailed report on areas for improvement and new proposals.

[0400] Step 8:

[0401] The system analyzes customer reviews and sentiment data in real time and makes product recommendations based on related topics and sentiment scores. The server extracts sentiment data from reviews and analyzes it in real time. The input is the review text data, and the output is product recommendations and related topics based on sentiment scores.

[0402] Step 9:

[0403] The emotion score distribution and clustering results are visualized on a dashboard. The server updates the dashboard so that users can check the emotion scores and clustering results in real time. The inputs are the emotion scores and clustering results, and the output is the visualized data on the dashboard.

[0404] By combining these processing steps, companies can efficiently collect and analyze huge amounts of data and take concrete measures to improve customer satisfaction.

[0405] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0406] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0407] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0408] [Second embodiment]

[0409] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0410] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0411] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0412] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0413] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0414] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0415] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0416] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0417] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0418] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0419] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0420] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0421] The present invention is a system that enables companies to efficiently process huge amounts of data and support future predictions and decision-making. The program processing of this system will be specifically described below.

[0422] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[0423] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[0424] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[0425] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[0426] The server visualizes the prediction and analysis results in a dashboard format and displays them on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[0427] Furthermore, the server analyzes customer feedback and behavioral data to suggest measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user, allowing the user to take appropriate actions to increase customer satisfaction.

[0428] Specific examples

[0429] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[0430] The server then analyzes customer feedback using NLP techniques, performing sentiment analysis and topic modeling, analyzing behavioral data and conducting clustering to identify customer purchasing patterns, which allows the server to identify desired features and shortcomings of new products.

[0431] The server trains a predictive model based on past data to forecast demand for new products. For example, it can estimate sales for the next three months. The forecast and analysis results are visualized in a dashboard format and displayed on the device.

[0432] Users can check the results through a dashboard and develop specific marketing strategies and production plans, taking into account the insights provided by the server. Furthermore, the server analyzes customer feedback and provides specific suggestions for improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[0433] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information due to the vast amount of data, make evidence-based decisions, and improve efficiency and competitiveness. It also contributes to improving customer satisfaction.

[0434] The processing flow will be explained below.

[0435] Step 1:

[0436] The server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server uses a data collection API to retrieve data from each source and stores it in a database.

[0437] Step 2:

[0438] The server cleans the data. The collected data may contain missing values ​​or outliers. The server detects these data and either fills in the missing values ​​with the average value or corrects the outliers. The server checks the data consistency to improve the data quality.

[0439] Step 3:

[0440] The server formats the data. To format the data in a form that is easy to analyze, it selects features and normalizes the data. In this step, the server extracts the necessary columns and standardizes the scale of the numerical data.

[0441] Step 4:

[0442] The server uses natural language processing technology to analyze the text data. For example, it tokenizes customer feedback and extracts important keywords. It also performs sentiment analysis and topic modeling to understand trends in customer opinions and emotions.

[0443] Step 5:

[0444] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify user segments and purchasing patterns. It uses clustering algorithms to identify groups of users with similar behavior.

[0445] Step 6:

[0446] The server trains the predictive model. It trains time series analysis models and regression analysis models based on past data to predict future market demand and customer behavior. In this step, data is trained and tested to evaluate the performance of the model.

[0447] Step 7:

[0448] The server performs the prediction execution, using the trained model to predict future events, such as estimating sales figures or customer intent after a new product launch.

[0449] Step 8:

[0450] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts to create an interactive dashboard.

[0451] Step 9:

[0452] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[0453] Step 10:

[0454] The server presents measures to improve customer satisfaction. It analyzes customer feedback and behavioral data and makes specific suggestions for improving products and services. Users can take appropriate actions based on the suggestions presented via their devices.

[0455] Example 1

[0456] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0457] Modern companies need to collect huge amounts of data from social media, customer review sites, and other sources, but it is difficult to efficiently organize and analyze that data and use it for future predictions and decision-making. Furthermore, there is a lack of technology to automatically generate specific proposals to improve customer satisfaction by utilizing customer feedback and behavioral data, so these issues must be resolved.

[0458] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0459] In this invention, the server includes: means for collecting information from data sources such as social media, customer review sites, and internal CRM systems; means for converting the collected information into JSON or CSV format and storing it in a database; means for cleaning the stored information using a library such as Pandas and formatting the data; means for analyzing the formatted data using natural language processing technology to extract important keywords, sentiment, and topics; means for the server to analyze behavioral data using techniques such as K-means clustering and identify user segments and purchasing patterns; means for training ARIMA models and regression analysis using past data to predict future market demand and customer behavior; means for visualizing the prediction and analysis results in dashboard format using Matplotlib or Plotly and displaying them on a terminal; and means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction. This enables efficient processing of massive amounts of data, supporting corporate decision-making, and improving customer satisfaction.

[0460] A "data source" is the original location or medium where information is stored, such as social media, customer review sites, or internal CRM systems.

[0461] "Gathering" refers to the process by which the server obtains the required information from the data sources.

[0462] "JSON" and "CSV" are standard data formats for storing collected data.

[0463] A "database" is a storage system that systematically organizes and stores information, allowing it to be accessed and managed efficiently.

[0464] "Cleaning" is the process of correcting or removing missing or outlier values ​​in data.

[0465] The "Pandas library" is a library for data processing and analysis in the Python language.

[0466] "Formatting" is the process of converting data into a form that is easier to analyze.

[0467] "Natural language processing technology" refers to the technology that enables computers to understand and process human language.

[0468] "Keywords" are important words and phrases extracted from the data.

[0469] "Sentiment analysis" is the process of identifying and classifying emotions and opinions contained within text data.

[0470] "Topic modeling" is a technology that automatically extracts topics from large amounts of text data.

[0471] "K-means clustering" is a machine learning algorithm for dividing data into several clusters (groups).

[0472] A "user segment" is a collection or group of users who share common characteristics.

[0473] "Purchasing patterns" are behaviors and tendencies that can be identified from a user's past purchasing history.

[0474] The "ARIMA model" is a statistical model used to analyze and predict time series data.

[0475] "Regression analysis" is a statistical method for filling in the relationships between variables.

[0476] "Forecasting" is the estimation of future events based on past data.

[0477] "Visualization" refers to the visual display of data and its analysis results using graphs and charts.

[0478] A "dashboard" is an interface for displaying multiple pieces of information on a single screen.

[0479] "Customer feedback" refers to opinions and evaluations from customers.

[0480] "Measures to improve customer satisfaction" are specific measures and proposals to increase customer satisfaction.

[0481] This invention is a system that enables companies to efficiently process large amounts of data and support future predictions and decision-making. This system is composed of a server, terminals, and users.

[0482] Server Processing

[0483] The server first collects information from data sources such as social media, customer review sites, and internal CRM systems. This information is collected using scripts and APIs written in Python. For example, to use the Twitter API to collect tweets containing a specific hashtag, the following method is used:

[0484] The server converts the collected data into JSON or CSV format and stores it in a database. This unifies the information collected from different data sources and makes subsequent processing easier. The server also uses the Pandas library to clean the data, correcting missing values ​​and outliers.

[0485] Data Shaping and Analysis

[0486] The server then formats the data and converts it into a format suitable for analysis. For example, it tokenizes the text data of customer reviews and removes stop words. Using this formatted data, the server analyzes customer feedback using natural language processing techniques. This processing involves keyword extraction using Scikit-learn's TfidfVectorizer and sentiment analysis using NLTK.

[0487] Clustering and Prediction

[0488] The server then uses algorithms such as K-means clustering to analyze behavioral data and identify user segments and purchasing patterns. It also uses historical data to train ARIMA models and regression analysis to predict future market demand and customer behavior. For example, an ARIMA model is used to forecast sales for the next three months based on past sales data.

[0489] Visualization and Presentation

[0490] The server visualizes the prediction and analysis results in dashboard format using Matplotlib or Plotly. This allows the data sent to the terminal to be displayed in a visually easy-to-understand format. Users can use this dashboard to check the analysis and prediction results and develop specific marketing strategies and production plans. The server also reanalyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction to users. This allows users to easily receive suggestions for product improvements and new services.

[0491] Specific examples

[0492] For example, consider a company planning to launch a new product. The server uses the Twitter API to collect tweets containing the hashtag "new product" and stores them in a database. The Pandas library is then used to impute missing values ​​in the collected data, tokenize it, and remove stop words. Next, Scikit-learn and NLTK are used to extract important keywords and perform sentiment analysis. K-means is used for clustering, and an ARIMA model is used to predict sales for the next three months. These results are visualized with Matplotlib and sent to the device.

[0493] Prompt Sentence Examples

[0494] "Write a program that uses the Twitter API to collect tweets containing a specific hashtag, cleans the data using Pandas, extracts important keywords using TfidfVectorizer, and then uses an ARIMA model to forecast sales for the next three months."

[0495] As a result, this system can efficiently process huge amounts of data, support corporate decision-making, and improve customer satisfaction.

[0496] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0497] Step 1:

[0498] The server collects data from social media, customer review sites, and internal CRM systems.

[0499] Input: Twitter API, customer review site API, data access method for internal CRM system

[0500] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "new product." Next, it collects product reviews from customer review sites via the API and retrieves customer purchase history data from the internal CRM system.

[0501] Output: Collected raw data (JSON and CSV formats)

[0502] Step 2:

[0503] The server stores the collected data in a database.

[0504] Input: Collected raw data (JSON or CSV format)

[0505] Specific operation: The server converts the collected data into JSON or CSV format and saves it in the corresponding table in the database. For example, Twitter data is saved in a table called "SNS_data".

[0506] Output: Data stored in a database in a uniform format

[0507] Step 3:

[0508] The server cleans the data.

[0509] Input: Data stored in a database

[0510] Specific operation: The server uses the Pandas library to detect, correct, or remove missing and outlier values. For example, in a customer review dataset, it imputes missing values ​​with the mean value and filters out outliers.

[0511] Output: Cleaned dataframe

[0512] Step 4:

[0513] The server formats the data and converts it into a format suitable for analysis.

[0514] Input: Cleaned dataframe

[0515] What it does: The server tokenizes the customer review text data and removes stop words. Specifically, it vectorizes the data using Scikit-learn's TfidfVectorizer.

[0516] Output: Tokenized text data with stop words removed

[0517] Step 5:

[0518] The server analyzes the data using natural language processing technology.

[0519] Input: Tokenized text data with stop words removed

[0520] How it works: The server extracts important keywords, performs sentiment analysis using NLTK, and extracts major topics using LDA topic modeling.

[0521] Output: Keyword list, sentiment score, topic list

[0522] Step 6:

[0523] The server clusters the behavioral data.

[0524] Input: Organized behavioral data

[0525] How it works: The server uses the K-means clustering algorithm to group users' purchasing patterns, for example, dividing users into multiple clusters based on purchase frequency or purchase amount.

[0526] Output: User's cluster information

[0527] Step 7:

[0528] The server uses historical data to train the predictive model.

[0529] Input: Existing sales data and customer behavior data

[0530] What it does: The server uses an ARIMA model to forecast sales for the next three months, and also trains the model to forecast market demand using regression analysis.

[0531] Output: Sales forecast results, demand forecast results

[0532] Step 8:

[0533] The server visualizes the prediction and analysis results.

[0534] Input: Sales forecast results, keyword list, sentiment score, user cluster information

[0535] Specific operation: The server uses Matplotlib or Plotly to visualize the prediction results as line graphs, pie charts, and bar graphs.

[0536] Output: Visualized graphs and charts

[0537] Step 9:

[0538] The terminal presents the visualized results to the user.

[0539] Input: Visualized graphs and charts

[0540] Specific operation: The terminal launches the dashboard and displays visualized data on the screen. Users can access the dashboard via a PC or tablet and check the analysis results.

[0541] Output: Analysis results displayed in a dashboard

[0542] Step 10:

[0543] The server then re-analyzes customer feedback and behavioral data to generate suggestions for improvements and new services.

[0544] Input: Customer feedback, behavioral data

[0545] How it works: The server analyzes negative customer feedback and uses machine learning algorithms to identify areas for product improvement. It also automatically generates new service proposals.

[0546] Output: List of improvement proposals, new service proposals

[0547] The above are the specific processing steps of this system.

[0548] (Application example 1)

[0549] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0550] In recent years, companies have been able to obtain vast amounts of data, but there is a demand for efficient systems that can effectively utilize this data and help with business decision-making. In particular, brick-and-mortar stores lack the means to quickly analyze data collected in real time and develop sales forecasts and marketing strategies. Furthermore, there is a demand for rapid response by providing store staff with analysis of customer feedback and demand forecasts in real time. To address these challenges, the present invention provides advanced data processing technology and a means for visualizing data in real time.

[0551] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0552] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and shaping the data, means for analyzing the shaped data using natural language processing technology and extracting important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction, means for cleaning the collected data in real time and performing analysis, sentiment analysis, and topic modeling using natural language processing technology, means for making demand forecasts based on past sales data and providing sales forecasts for the next period, and means for displaying data visualizations on smart glasses and notifying important trends and alerts in real time. This enables companies to extract useful information from massive amounts of data and use it for customer analysis, demand forecasting, and sales management.

[0553] A "data source" is the source of the original data used to collect information, and includes social media sites and customer review sites.

[0554] "Means for collecting information" refers to the techniques and methods for efficiently acquiring the necessary information from data sources.

[0555] A "database" is a collection of structured data that is used to systematically store and manage collected information.

[0556] "Information cleaning methods" are techniques and methods for correcting missing data and outliers from collected data and improving the quality of the data.

[0557] "Means for data preparation" refers to techniques and methods for converting cleaned data into an analyzable form and formatting it appropriately.

[0558] "Natural language processing technology" is a computer technology for processing human language, including text tokenization, sentiment analysis, and topic modeling.

[0559] "Means for extracting important keywords and patterns" refers to techniques and methods for finding important information and trends that are useful for business from analyzed data.

[0560] A "means for training a predictive model" is a technique or method for building and training a model to predict future events or conditions based on past data.

[0561] "Means for making future predictions" are technologies or methods that use trained predictive models to forecast future demand, sales, etc.

[0562] "Means for visualizing in dashboard format" refers to techniques and methods for displaying analysis results and prediction results in a format that is intuitively easy to understand, including graphs and charts.

[0563] "Means for displaying on a device" refers to the technology or method for displaying the information visualized in dashboard format on a device used by the user (e.g., smart glasses, tablet).

[0564] "Means for analyzing customer feedback and behavioral data" refers to techniques and methods for analyzing customer opinions and behavioral history to evaluate customer satisfaction and behavioral trends.

[0565] "Means for proposing measures to improve customer satisfaction" refers to techniques and methods that propose specific actions to improve customer satisfaction based on customer feedback and behavioral data.

[0566] "Means for cleaning data collected in real time" refers to techniques and methods for immediately correcting missing or outliers in data collected in real time, thereby maintaining the quality of the data.

[0567] "Means for sentiment analysis and topic modeling" refers to techniques and methods that use natural language processing techniques to analyze sentiment trends and major topics from text.

[0568] "Means for forecasting demand based on past sales data" refers to techniques and methods for forecasting future sales and demand using historical sales data.

[0569] The "means for providing a sales forecast for the next period" refers to a technique or method for predicting future sales for a specific period and providing the results.

[0570] "Means for displaying on smart glasses" means the technology or method for displaying visualized data or alerts on smart glasses, allowing users to view the information in real time.

[0571] "Means for notifying important trends and alerts" refers to technologies and methods that immediately notify users of trends or anomalies that require particular attention based on the results of data analysis.

[0572] The "data analysis system for brick-and-mortar stores" of the present invention effectively collects huge amounts of data and analyzes it in real time to optimize brick-and-mortar store operations. Specific embodiments of this system are described below.

[0573] 1. Data Collection

[0574] The server collects customer feedback and sales data from social media, customer review sites, internal CRM systems, etc. For example, it uses the Twitter API to retrieve tweets related to product names and collect customer opinions.

[0575] 2. Data cleaning and formatting

[0576] The collected data is cleaned using the Pandas library to correct missing data and outliers and improve data quality. This cleaned data is then formatted using natural language processing techniques, specifically NLTK and SpaCy, to tokenize the text and convert it into a parseable format.

[0577] 3. Data analysis

[0578] The server uses natural language processing technology to perform sentiment analysis and topic modeling on the formatted data. For example, it analyzes customer tweets and determines whether their content is positive or negative. This allows for a quantitative understanding of product reviews on social media.

[0579] 4. Demand forecasting

[0580] Based on past sales data, the server uses Scikit-learn to train a predictive model. For example, monthly sales data is used to forecast sales for the next three months. The trained model can then predict future demand with high accuracy.

[0581] 5. Data Visualization and Notification

[0582] The server visualizes the prediction and analysis results in the form of a dashboard. Graphs and charts are created using Matplotlib and Seaborn. The generated visualization data is displayed in real time on smart glasses via AR Kit or the Google Glass API. This allows users to receive the information they need to run their stores in real time and instantly grasp important trends and alerts.

[0583] Specific examples

[0584] As a concrete example, consider a store planning to launch a new product. In this system, the server collects information about the new product from social media and customer review sites and performs sentiment analysis. For example, if a prompt such as "Collect social media reactions about product X and display the results of the sentiment analysis on the smart glasses in real time" is input into the generative AI model, the results of the sentiment analysis will be displayed in real time on the smart glasses. In addition, sales forecasts for the next three months are generated based on past sales data and notified to the user. This can be used to develop marketing strategies and production plans for the new product.

[0585] The present invention is a system that eliminates the difficulty of understanding information due to the vast amount of data and provides the ability to make evidence-based decisions, thereby significantly improving the competitiveness of physical stores.

[0586] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0587] Processing Steps

[0588] Step 1:

[0589] The server collects information from data sources (social media, customer review sites, internal CRM systems, etc.).

[0590] Input: API key or authentication information for data collection, search keywords.

[0591] Output: The raw data collected.

[0592] Specific operation: The server uses the Twitter API to search for tweets related to "product name" and obtains the results in text format.

[0593] Step 2:

[0594] The server stores the collected raw data in a database.

[0595] Input: Collected raw data (text data).

[0596] Output: The saved database entries.

[0597] Specific operation: The server converts the tweet data into a tabular format using a Pandas data frame and saves it in an SQL database.

[0598] Step 3:

[0599] The server cleans the stored data and corrects missing or outlier values.

[0600] Input: A saved database entry.

[0601] Output: Cleaned data.

[0602] Specific operation: The server uses the Pandas library to remove or impute missing data and correct outliers.

[0603] Step 4:

[0604] The server formats the cleaned data and applies natural language processing techniques.

[0605] Input: Cleaned data.

[0606] Output: Formatted text data.

[0607] Specific operation: The server uses the NLTK or SpaCy library to tokenize the text data, remove stop words, and extract the necessary features.

[0608] Step 5:

[0609] The server analyzes the data using natural language processing techniques to perform sentiment analysis and topic modeling.

[0610] Input: Formatted text data.

[0611] Output: Analysis results (sentiment scores, topic distribution).

[0612] Specific operation: The server performs sentiment analysis of the text using NLTK's SentimentIntensityAnalyzer and extracts important topics using an LDA model.

[0613] Step 6:

[0614] The server trains a predictive model based on past sales data and makes future predictions.

[0615] Input: Historical sales data, cleaned data.

[0616] Output: Sales forecast results.

[0617] What happens: The server uses Scikit-learn to train a linear regression model and perform sales forecasts for the next three months.

[0618] Step 7:

[0619] The server visualizes the prediction and analysis results in dashboard format and displays them on the terminal.

[0620] Input: Sales forecast results, sentiment analysis and topic modeling results.

[0621] Output: Visualized graphs on a dashboard.

[0622] Specific operation: The server creates graphs and charts using Matplotlib and Seaborn and generates a web dashboard.

[0623] Step 8:

[0624] The server analyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction.

[0625] Input: Customer feedback data, behavioral data.

[0626] Output: Specific action proposals to improve customer satisfaction.

[0627] Specific operation: The server performs sentiment analysis based on the feedback data, extracts areas for improvement in products and services, and notifies the marketing team.

[0628] Step 9:

[0629] The server cleans the collected data in real time and uses natural language processing technology to perform analysis, sentiment analysis and topic modeling.

[0630] Input: Raw data collected in real time.

[0631] Output: Cleaned and analyzed data.

[0632] Specific operation: The server cleans real-time streaming data with Pandas, analyzes sentiment scores with NLTK, and extracts topics with an LDA model.

[0633] Step 10:

[0634] The server makes a demand forecast based on past sales data and provides a sales forecast for the next period.

[0635] Input: Historical sales data.

[0636] Output: Sales forecast for next period.

[0637] What it does: The server uses Scikit-learn's time series forecasting model to forecast future sales and generate results.

[0638] Step 11:

[0639] The server displays data visualizations on the smart glasses, notifying users of important trends and alerts in real time.

[0640] Inputs: Visualization data, important trends or alerts.

[0641] Output: Real-time data displayed on smart glasses.

[0642] What it does: The server uses AR Kit or Google Glass APIs to display important data on the smart glasses and notify you of trends and alerts.

[0643] Through the above steps, this system can effectively collect and analyze huge amounts of data, significantly supporting decision-making in brick-and-mortar store operations.

[0644] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0645] The present invention is a system that aims to enable companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, making it possible to present more sophisticated measures to improve customer satisfaction. The program processing of this system is explained in detail below.

[0646] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[0647] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[0648] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[0649] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[0650] The server visualizes the prediction and analysis results in the form of a dashboard and displays it on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[0651] Furthermore, in this invention, the server analyzes customer feedback and behavioral data to present measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user. Here, the emotion engine plays a particularly important role. The emotion engine recognizes the user's emotions and stores the emotion data in a database. The server also analyzes this emotion data to assist in providing more precise measures to improve customer satisfaction.

[0652] Specific examples

[0653] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[0654] The server uses NLP techniques to analyze customer feedback, perform sentiment analysis and topic modeling, analyze behavioral data, and perform clustering to identify customer buying patterns, which allows the server to identify desired features and shortcomings of new products.

[0655] The server then trains a predictive model, forecasting demand for new products based on past data and estimating sales figures for the next three months, for example. The forecast and analysis results are then visualized in a dashboard format and displayed on the device.

[0656] Users can check the results through the dashboard and develop specific marketing strategies and production plans, while also referring to the data from the emotion engine provided by the server. The emotion engine analyzes customer emotions, and the server uses this data to suggest specific improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[0657] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information caused by vast amounts of data, make evidence-based decisions, and improve efficiency and competitiveness.In addition, by utilizing the emotion engine, it can further contribute to improving customer satisfaction.

[0658] The processing flow will be explained below.

[0659] Step 1:

[0660] The server collects information from data sources, such as social media, customer review sites, and internal CRM systems. The server uses data collection APIs to retrieve data from each source and stores it in a database.

[0661] Step 2:

[0662] The server cleans the stored data, removing missing data and noise, and detecting and correcting outliers, for example by imputing missing values ​​with the mean or correcting outliers to improve the quality of the data.

[0663] Step 3:

[0664] The server formats the data, converts it into a format that is easy to analyze, extracts the necessary features, and normalizes the data so that all data is handled on a uniform scale.

[0665] Step 4:

[0666] The server uses natural language processing (NLP) to analyze the text data, tokenize customer feedback, extract key keywords and phrases, and perform sentiment analysis and topic modeling to understand trends in customer opinions and sentiment.

[0667] Step 5:

[0668] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify patterns in user behavior. It uses a clustering algorithm to group users with similar behavior.

[0669] Step 6:

[0670] The server uses an emotion engine to recognize the user's emotions. For example, it analyzes emotions from customer text feedback and facial expression data. The emotion data is stored in a database and integrated with other data for analysis.

[0671] Step 7:

[0672] The server trains the predictive model. Based on past data, it uses time series analysis and regression analysis models to predict future market demand and customer behavior. The predictive model is then evaluated for accuracy through training and testing on data.

[0673] Step 8:

[0674] The server performs the prediction execution, using the trained model to make specific predictions about future events, such as estimating sales figures or customer intent after a new product launch.

[0675] Step 9:

[0676] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts and displays the information on an interactive dashboard.

[0677] Step 10:

[0678] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[0679] Step 11:

[0680] The server proposes measures to improve customer satisfaction based on data from the emotion engine. It comprehensively analyzes customer feedback, behavioral data, and recognized emotional data to propose specific improvements to products and services. The proposals are presented via the device, and the user can then take action to increase customer satisfaction.

[0681] Example 2

[0682] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0683] Modern companies face challenges in efficiently collecting massive amounts of data, analyzing it, and making future predictions and decisions. Furthermore, a lack of precise ways to recognize customer sentiment and improve customer satisfaction makes it difficult to properly understand customer needs and respond quickly. Existing systems perform data cleaning, natural language processing analysis, and predictive model training separately, making integrated analysis and prediction difficult, and making effective decision-making difficult.

[0684] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0685] In this invention, the server includes: [means for collecting information from data sources;] [means for storing the collected information in a database;] [means for cleaning the stored information and formatting the data;] [means for analyzing the formatted data using natural language processing technology and extracting important keywords and patterns;] [means for training a predictive model based on past data and making future predictions;] [means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal; and [means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction using an emotion engine.] This enables companies to collect and analyze data in an integrated manner, make future predictions and decisions, and quickly and effectively respond to customer needs and improve customer satisfaction.

[0686] "Data sources" are sources of data collected from external and internal systems and platforms, such as social media, customer review sites, and CRM systems.

[0687] A "database" is an information system that stores collected information and enables it to be efficiently managed and searched. Examples include relational databases and NoSQL databases.

[0688] "Cleaning" is the process of correcting and removing missing data and outliers from stored information to improve data quality.

[0689] "Natural language processing technology" refers to computer techniques for analyzing, understanding, and generating human language. Specifically, it includes tokenization, sentiment analysis, and topic modeling.

[0690] A "predictive model" is a statistical or machine learning model trained to predict future trends based on past data. It includes techniques such as time series analysis and regression analysis.

[0691] A "dashboard" is an interface that visually displays analysis and prediction results, allowing users to intuitively grasp the situation.

[0692] An "emotion engine" is a technology or algorithm that analyzes user emotions and generates corresponding emotional data. It helps to precisely improve customer satisfaction.

[0693] "Customer satisfaction improvement measures" are specific action plans for presenting improvements and new proposals for products and services based on customer evaluations and feedback.

[0694] The present invention is a system that enables companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, by incorporating an emotion engine that recognizes user emotions, it is possible to provide more sophisticated measures to improve customer satisfaction. Specific embodiments of this system are described below.

[0695] First, the server collects information from data sources. Data sources include social media, customer review sites, and internal CRM systems. For example, the server might use an API to collect data related to a specific hashtag from a social media site. The server then converts the collected information into an appropriate format and stores it in a database. Relational databases or NoSQL databases are often used for this purpose.

[0696] As a concrete example, a server uses the Twitter API to collect tweets containing the hashtag "new product release," converts them into JSON format, and stores them in a database.

[0697] Next, the server cleans the stored information. This step uses Python's Pandas library to impute missing data and correct outliers. For example, records with missing purchase dates are imputed with the user's closest purchase date. Additionally, any outliers that exist are removed through manual review or automated correction.

[0698] The server then formats the data, selects and normalizes the necessary features, for example, converting customer age, purchase frequency, and purchase amount to a standardized scale using scaling and feature selection techniques. The formatted data is then split into a training dataset and a test dataset.

[0699] The server then uses natural language processing (NLP) techniques to analyze this data. Specifically, it uses the SpaCy library to tokenize customer feedback and extract key keywords. It also performs sentiment analysis to assign a positive, negative, or neutral sentiment score to the feedback. Finally, it uses Scikit-learn to cluster behavioral data and identify customer purchasing patterns.

[0700] Next, the server uses TensorFlow to train a predictive model for time series analysis. For example, it predicts sales figures for the next three months based on past data. The predictive model sets evaluation indicators and evaluates reliability. If anomalies are detected, the model can be retrained.

[0701] The server uses Tableau to visualize these predictions and analysis results in dashboard format and displays them on the device. Users can access the dashboard using a PC or tablet and intuitively grasp the situation, which is useful for formulating marketing strategies and production plans.

[0702] Furthermore, the server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The emotion engine analyzes the user's emotions and stores the results in a database. The server then generates measures to improve customer satisfaction based on the emotion data and proposes them to the user.

[0703] Examples of prompt statements

[0704] How do I impute missing data and remove outliers using Python's Pandas?

[0705] "Describe an effective method for conducting sentiment analysis from customer feedback using natural language processing."

[0706] "How do I train a time series analysis model using TensorFlow to forecast sales for the next three months?"

[0707] "Please explain how to visualize data collected with Tableau in the form of a dashboard."

[0708] This allows companies to overcome the difficulty of comprehending information caused by the sheer volume of data, make evidence-based decisions, and quickly implement concrete actions to improve customer satisfaction.

[0709] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0710] Step 1:

[0711] The server collects information from data sources, including social media, customer review sites, and internal CRM systems. It retrieves data as input by sending API requests. Specifically, it uses the Twitter API to collect tweets containing the hashtag "new product release" and converts the retrieved data into JSON format. It generates a JSON object of the collected data as output.

[0712] Step 2:

[0713] The server stores the collected information in a database. The input is the JSON object collected in step 1. Specifically, it executes an INSERT query against an SQL or NoSQL database to store the data. The output is the number of records stored in the database.

[0714] Step 3:

[0715] The server cleans the stored information. The input is raw data read from the database. Specifically, it uses Python's Pandas library to impute missing data and correct outliers. For example, it imputes missing purchase dates with the user's closest purchase date. The output is a cleaned data frame.

[0716] Step 4:

[0717] The server formats the cleaned data. The input is a cleaned data frame. Specifically, features such as customer age, purchase frequency, and purchase amount are selected and normalized. Feature scaling is performed using Scikit-learn's StandardScaler. The output is a formatted data frame.

[0718] Step 5:

[0719] The server uses natural language processing techniques to analyze the formatted data. The input is a formatted data frame. Specifically, the SpaCy library is used to tokenize the customer feedback and extract important keywords. A sentiment analysis engine is used to calculate a sentiment score for the feedback. The output is a new data frame containing the analysis results.

[0720] Step 6:

[0721] The server clusters the behavioral data. The input is a data frame analyzed using natural language processing techniques. Specifically, it uses Scikit-learn to perform K-means clustering to identify customer purchasing patterns. The output is the data points corresponding to each cluster and the coordinates of the cluster center.

[0722] Step 7:

[0723] The server trains a predictive model. The input is clustered historical data. TensorFlow is used to train a predictive model for time series analysis. For example, a model is created to predict sales for the next three months. The output is a dataset containing the predicted values.

[0724] Step 8:

[0725] The server visualizes the prediction and analysis results in the form of a dashboard and presents it to the terminal. The input is a dataset containing the prediction values ​​and analysis results. Specifically, the dashboard is created using Tableau and provided to the end user through a web-based interface. The output is a visualized dashboard that can be viewed on the terminal.

[0726] Step 9:

[0727] The server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The input is customer feedback data. Specifically, the emotion engine analyzes user feedback and generates an emotion score. This is then stored in a database, and measures to improve customer satisfaction are generated based on the analysis results. The output is a report proposing measures to improve customer satisfaction.

[0728] Through these steps, the system enables businesses to efficiently collect and analyze data and provide actionable insights, which in turn contributes to improving customer satisfaction.

[0729] (Application example 2)

[0730] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0731] In recent years, companies have been required to quickly and accurately collect and analyze massive amounts of data and make appropriate decisions based on that data. However, due to the sheer volume and diversity of data, there is a lack of concrete methods for how to collect, analyze, and utilize this information. Furthermore, while there is a need to further improve customer satisfaction by recognizing customer emotions in real time, current systems have difficulty doing this efficiently. This creates the challenge of being unable to grasp the detailed needs and emotions of customers, making it difficult to present appropriate product proposals and improvement measures.

[0732] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0733] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and formatting the data, means for analyzing the formatted data using natural language processing technology to extract important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, means for analyzing customer reviews and emotion data in real time and recommending products based on related topics and emotion scores, and means for visualizing emotion score distributions and clustering results on the dashboard. This enables companies to efficiently collect and analyze massive amounts of data, recognize customer emotions in real time, and propose product suggestions and improvement measures.

[0734] A "data source" is a data provider that includes a file, database, or other source of information.

[0735] "Means for collecting information" refers to the tools and techniques used to obtain the necessary data from designated data sources and input it into the system.

[0736] "Means for storing information in a database" refers to the technology or system for storing collected data in an appropriate format as structured data.

[0737] "Information cleaning and data formatting methods" are processes and techniques that correct incomplete or erroneous data and convert it into a coherent form.

[0738] "Natural language processing technology" is a technology that allows computers to understand, process, and analyze human language.

[0739] A "predictive model" is a mathematical or statistical model used to predict future behavior or outcomes based on past data.

[0740] A "dashboard format" is an interface or layout that displays data in a visually easy-to-understand way.

[0741] "Customer feedback" refers to customer evaluations and opinions about products and services.

[0742] "Customer behavioral data" refers to data related to customer behavior, such as purchasing history and website browsing history.

[0743] "Customer reviews" are ratings and comments written by customers about products or services.

[0744] "Emotional data" is emotional information extracted from customer reviews and feedback.

[0745] "Real-time analysis" is the process of analyzing data immediately at the moment it is generated.

[0746] "Relevant topics" are the main discussion themes or topics that emerge from the analyzed data.

[0747] An "emotion score" is a numerical representation of the degree of positivity or negativity of emotions extracted from text data.

[0748] "Clustering" is a technique for grouping data based on similarity.

[0749] The present invention is a system for businesses to efficiently collect and analyze data and improve customer satisfaction. The system incorporates data collection, cleaning, natural language processing, predictive model training, visualization with dashboards, sentiment analysis, and clustering functions.

[0750] System Configuration

[0751] 1. Data Collection and Storage

[0752] The server collects data from multiple data sources (e.g., social media, customer review sites, CRM systems, etc.) and stores it in a database. This data collection is done using the requests library.

[0753] 2. Data cleaning and formatting

[0754] The server cleans and formats the collected data. This step involves correcting missing data and outliers to improve data quality. This is done using the Python standard library and Pandas.

[0755] 3. Natural Language Processing and Sentiment Analysis

[0756] The server analyzes the cleaned data using natural language processing (NLP) techniques. Specifically, it performs sentiment analysis using TextBlob to extract important keywords and patterns. It also uses KMeans from scikit-learn for clustering.

[0757] 4. Training a predictive model and forecasting the future

[0758] The server uses historical data to train predictive models that forecast future market demand and customer behavior, using algorithms such as time series analysis and regression analysis.

[0759] 5. Visualizing and Presenting Results

[0760] The prediction and analysis results are visualized in a dashboard format. Users can access the dashboard and check the results using a PC or tablet. Visualization is performed using libraries such as Matplotlib and Dash.

[0761] 6. Proposing measures to improve customer satisfaction

[0762] The server analyzes customer feedback and behavioral data to suggest satisfaction improvement measures tailored to each individual customer. The emotion engine analyzes customer reviews and emotion data in real time and recommends products based on related topics and emotion scores.

[0763] Specific examples

[0764] For example, when a user of the "Online Shopping Improvement Assistant" on the online shopping site views a specific product, such as a "Bluetooth speaker," the server analyzes the reviews on that page. If there are many negative sentiments based on sentiment analysis, the server automatically notifies the user about areas that need improvement. Related topics (e.g., sound quality, battery life, connectivity) extracted from the reviews through clustering are also presented.

[0765] Prompt Sentence Examples

[0766] "Write a Python program that extracts negative sentiment from Bluetooth speaker reviews and notifies users with related topics. Use the TextBlob and scikit-learn libraries."

[0767] These systems enable companies to effectively utilize vast amounts of data and provide specific product suggestions and improvement measures based on customer sentiment, while users can check the analysis results in real time and make decisions quickly.

[0768] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0769] Program processing steps

[0770] Step 1:

[0771] The server collects information from data sources. Specifically, it collects data from multiple data sources (such as social media, customer review sites, and internal CRM systems) through APIs. It primarily uses the requests library to retrieve data in JSON format. The input is the URL and API key of the data source, and the output is the collected raw data.

[0772] Step 2:

[0773] Store the collected information in a database. The server converts the raw data into the required format and stores it in an appropriate database (e.g., MySQL or MongoDB). The input is the collected raw data, and the output is structured data stored in the database.

[0774] Step 3:

[0775] The stored information is cleaned and the data is formatted. The server uses the Pandas library to correct incomplete data and outliers and standardize the data format. The input is raw data retrieved from the database, and the output is cleaned, high-quality data.

[0776] Step 4:

[0777] The formatted data is analyzed using natural language processing techniques. The server uses the TextBlob library to analyze the text data of each review and calculate a sentiment score. It also performs clustering using scikit-learn. The input is the cleaned data, and the output is the sentiment score and clustering results for each review.

[0778] Step 5:

[0779] The server trains a predictive model based on historical data to predict market demand and customer behavior using a predictive algorithm. The input is the training data, and the output is the trained predictive model.

[0780] Step 6:

[0781] The prediction and analysis results are visualized in dashboard format and displayed on the terminal. The server uses libraries such as Matplotlib and Dash to visualize the sentiment scores, clustering results, and prediction results in graphs and charts. The input is the prediction model and analysis data, and the output is a visually easy-to-understand dashboard.

[0782] Step 7:

[0783] Customer feedback and behavioral data are analyzed to suggest measures to improve customer satisfaction. The server generates a specific action plan to improve customer satisfaction based on the analysis results and presents it to the user. The input is feedback data and behavioral data, and the output is a detailed report on areas for improvement and new proposals.

[0784] Step 8:

[0785] The system analyzes customer reviews and sentiment data in real time and makes product recommendations based on related topics and sentiment scores. The server extracts sentiment data from reviews and analyzes it in real time. The input is the review text data, and the output is product recommendations and related topics based on sentiment scores.

[0786] Step 9:

[0787] The emotion score distribution and clustering results are visualized on a dashboard. The server updates the dashboard so that users can check the emotion scores and clustering results in real time. The inputs are the emotion scores and clustering results, and the output is the visualized data on the dashboard.

[0788] By combining these processing steps, companies can efficiently collect and analyze huge amounts of data and take concrete measures to improve customer satisfaction.

[0789] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0790] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0791] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0792] [Third embodiment]

[0793] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0794] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0795] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0796] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0797] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0798] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0799] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0800] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0801] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0802] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0803] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0804] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0805] The present invention is a system that enables companies to efficiently process huge amounts of data and support future predictions and decision-making. The program processing of this system will be specifically described below.

[0806] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[0807] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[0808] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[0809] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[0810] The server visualizes the prediction and analysis results in a dashboard format and displays them on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[0811] Furthermore, the server analyzes customer feedback and behavioral data to suggest measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user, allowing the user to take appropriate actions to increase customer satisfaction.

[0812] Specific examples

[0813] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[0814] The server then analyzes customer feedback using NLP techniques, performing sentiment analysis and topic modeling, analyzing behavioral data and conducting clustering to identify customer purchasing patterns, which allows the server to identify desired features and shortcomings of new products.

[0815] The server trains a predictive model based on past data to forecast demand for new products. For example, it can estimate sales for the next three months. The forecast and analysis results are visualized in a dashboard format and displayed on the device.

[0816] Users can check the results through a dashboard and develop specific marketing strategies and production plans, taking into account the insights provided by the server. Furthermore, the server analyzes customer feedback and provides specific suggestions for improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[0817] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information due to the vast amount of data, make evidence-based decisions, and improve efficiency and competitiveness. It also contributes to improving customer satisfaction.

[0818] The processing flow will be explained below.

[0819] Step 1:

[0820] The server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server uses a data collection API to retrieve data from each source and stores it in a database.

[0821] Step 2:

[0822] The server cleans the data. The collected data may contain missing values ​​or outliers. The server detects these data and either fills in the missing values ​​with the average value or corrects the outliers. The server checks the data consistency to improve the data quality.

[0823] Step 3:

[0824] The server formats the data. To format the data in a form that is easy to analyze, it selects features and normalizes the data. In this step, the server extracts the necessary columns and standardizes the scale of the numerical data.

[0825] Step 4:

[0826] The server uses natural language processing technology to analyze the text data. For example, it tokenizes customer feedback and extracts important keywords. It also performs sentiment analysis and topic modeling to understand trends in customer opinions and emotions.

[0827] Step 5:

[0828] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify user segments and purchasing patterns. It uses clustering algorithms to identify groups of users with similar behavior.

[0829] Step 6:

[0830] The server trains the predictive model. It trains time series analysis models and regression analysis models based on past data to predict future market demand and customer behavior. In this step, data is trained and tested to evaluate the performance of the model.

[0831] Step 7:

[0832] The server performs the prediction execution, using the trained model to predict future events, such as estimating sales figures or customer intent after a new product launch.

[0833] Step 8:

[0834] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts to create an interactive dashboard.

[0835] Step 9:

[0836] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[0837] Step 10:

[0838] The server presents measures to improve customer satisfaction. It analyzes customer feedback and behavioral data and makes specific suggestions for improving products and services. Users can take appropriate actions based on the suggestions presented via their devices.

[0839] Example 1

[0840] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0841] Modern companies need to collect huge amounts of data from social media, customer review sites, and other sources, but it is difficult to efficiently organize and analyze that data and use it for future predictions and decision-making. Furthermore, there is a lack of technology to automatically generate specific proposals to improve customer satisfaction by utilizing customer feedback and behavioral data, so these issues must be resolved.

[0842] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0843] In this invention, the server includes: means for collecting information from data sources such as social media, customer review sites, and internal CRM systems; means for converting the collected information into JSON or CSV format and storing it in a database; means for cleaning the stored information using a library such as Pandas and formatting the data; means for analyzing the formatted data using natural language processing technology to extract important keywords, sentiment, and topics; means for the server to analyze behavioral data using techniques such as K-means clustering and identify user segments and purchasing patterns; means for training ARIMA models and regression analysis using past data to predict future market demand and customer behavior; means for visualizing the prediction and analysis results in dashboard format using Matplotlib or Plotly and displaying them on a terminal; and means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction. This enables efficient processing of massive amounts of data, supporting corporate decision-making, and improving customer satisfaction.

[0844] A "data source" is the original location or medium where information is stored, such as social media, customer review sites, or internal CRM systems.

[0845] "Gathering" refers to the process by which the server obtains the required information from the data sources.

[0846] "JSON" and "CSV" are standard data formats for storing collected data.

[0847] A "database" is a storage system that systematically organizes and stores information, allowing it to be accessed and managed efficiently.

[0848] "Cleaning" is the process of correcting or removing missing or outlier values ​​in data.

[0849] The "Pandas library" is a library for data processing and analysis in the Python language.

[0850] "Formatting" is the process of converting data into a form that is easier to analyze.

[0851] "Natural language processing technology" refers to the technology that enables computers to understand and process human language.

[0852] "Keywords" are important words and phrases extracted from the data.

[0853] "Sentiment analysis" is the process of identifying and classifying emotions and opinions contained within text data.

[0854] "Topic modeling" is a technology that automatically extracts topics from large amounts of text data.

[0855] "K-means clustering" is a machine learning algorithm for dividing data into several clusters (groups).

[0856] A "user segment" is a collection or group of users who share common characteristics.

[0857] "Purchasing patterns" are behaviors and tendencies that can be identified from a user's past purchasing history.

[0858] The "ARIMA model" is a statistical model used to analyze and predict time series data.

[0859] "Regression analysis" is a statistical method for filling in the relationships between variables.

[0860] "Forecasting" is the estimation of future events based on past data.

[0861] "Visualization" refers to the visual display of data and its analysis results using graphs and charts.

[0862] A "dashboard" is an interface for displaying multiple pieces of information on a single screen.

[0863] "Customer feedback" refers to opinions and evaluations from customers.

[0864] "Measures to improve customer satisfaction" are specific measures and proposals to increase customer satisfaction.

[0865] This invention is a system that enables companies to efficiently process large amounts of data and support future predictions and decision-making. This system is composed of a server, terminals, and users.

[0866] Server Processing

[0867] The server first collects information from data sources such as social media, customer review sites, and internal CRM systems. This information is collected using scripts and APIs written in Python. For example, to use the Twitter API to collect tweets containing a specific hashtag, the following method is used:

[0868] The server converts the collected data into JSON or CSV format and stores it in a database. This unifies the information collected from different data sources and makes subsequent processing easier. The server also uses the Pandas library to clean the data, correcting missing values ​​and outliers.

[0869] Data Shaping and Analysis

[0870] The server then formats the data and converts it into a format suitable for analysis. For example, it tokenizes the text data of customer reviews and removes stop words. Using this formatted data, the server analyzes customer feedback using natural language processing techniques. This processing involves keyword extraction using Scikit-learn's TfidfVectorizer and sentiment analysis using NLTK.

[0871] Clustering and Prediction

[0872] The server then uses algorithms such as K-means clustering to analyze behavioral data and identify user segments and purchasing patterns. It also uses historical data to train ARIMA models and regression analysis to predict future market demand and customer behavior. For example, an ARIMA model is used to forecast sales for the next three months based on past sales data.

[0873] Visualization and Presentation

[0874] The server visualizes the prediction and analysis results in dashboard format using Matplotlib or Plotly. This allows the data sent to the terminal to be displayed in a visually easy-to-understand format. Users can use this dashboard to check the analysis and prediction results and develop specific marketing strategies and production plans. The server also reanalyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction to users. This allows users to easily receive suggestions for product improvements and new services.

[0875] Specific examples

[0876] For example, consider a company planning to launch a new product. The server uses the Twitter API to collect tweets containing the hashtag "new product" and stores them in a database. The Pandas library is then used to impute missing values ​​in the collected data, tokenize it, and remove stop words. Next, Scikit-learn and NLTK are used to extract important keywords and perform sentiment analysis. K-means is used for clustering, and an ARIMA model is used to predict sales for the next three months. These results are visualized with Matplotlib and sent to the device.

[0877] Prompt Sentence Examples

[0878] "Write a program that uses the Twitter API to collect tweets containing a specific hashtag, cleans the data using Pandas, extracts important keywords using TfidfVectorizer, and then uses an ARIMA model to forecast sales for the next three months."

[0879] As a result, this system can efficiently process huge amounts of data, support corporate decision-making, and improve customer satisfaction.

[0880] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0881] Step 1:

[0882] The server collects data from social media, customer review sites, and internal CRM systems.

[0883] Input: Twitter API, customer review site API, data access method for internal CRM system

[0884] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "new product." Next, it collects product reviews from customer review sites via the API and retrieves customer purchase history data from the internal CRM system.

[0885] Output: Collected raw data (JSON and CSV formats)

[0886] Step 2:

[0887] The server stores the collected data in a database.

[0888] Input: Collected raw data (JSON or CSV format)

[0889] Specific operation: The server converts the collected data into JSON or CSV format and saves it in the corresponding table in the database. For example, Twitter data is saved in a table called "SNS_data".

[0890] Output: Data stored in a database in a uniform format

[0891] Step 3:

[0892] The server cleans the data.

[0893] Input: Data stored in a database

[0894] Specific operation: The server uses the Pandas library to detect, correct, or remove missing and outlier values. For example, in a customer review dataset, it imputes missing values ​​with the mean value and filters out outliers.

[0895] Output: Cleaned dataframe

[0896] Step 4:

[0897] The server formats the data and converts it into a format suitable for analysis.

[0898] Input: Cleaned dataframe

[0899] What it does: The server tokenizes the customer review text data and removes stop words. Specifically, it vectorizes the data using Scikit-learn's TfidfVectorizer.

[0900] Output: Tokenized text data with stop words removed

[0901] Step 5:

[0902] The server analyzes the data using natural language processing technology.

[0903] Input: Tokenized text data with stop words removed

[0904] How it works: The server extracts important keywords, performs sentiment analysis using NLTK, and extracts major topics using LDA topic modeling.

[0905] Output: Keyword list, sentiment score, topic list

[0906] Step 6:

[0907] The server clusters the behavioral data.

[0908] Input: Organized behavioral data

[0909] How it works: The server uses the K-means clustering algorithm to group users' purchasing patterns, for example, dividing users into multiple clusters based on purchase frequency or purchase amount.

[0910] Output: User's cluster information

[0911] Step 7:

[0912] The server uses historical data to train the predictive model.

[0913] Input: Existing sales data and customer behavior data

[0914] What it does: The server uses an ARIMA model to forecast sales for the next three months, and also trains the model to forecast market demand using regression analysis.

[0915] Output: Sales forecast results, demand forecast results

[0916] Step 8:

[0917] The server visualizes the prediction and analysis results.

[0918] Input: Sales forecast results, keyword list, sentiment score, user cluster information

[0919] Specific operation: The server uses Matplotlib or Plotly to visualize the prediction results as line graphs, pie charts, and bar graphs.

[0920] Output: Visualized graphs and charts

[0921] Step 9:

[0922] The terminal presents the visualized results to the user.

[0923] Input: Visualized graphs and charts

[0924] Specific operation: The terminal launches the dashboard and displays visualized data on the screen. Users can access the dashboard via a PC or tablet and check the analysis results.

[0925] Output: Analysis results displayed in a dashboard

[0926] Step 10:

[0927] The server then re-analyzes customer feedback and behavioral data to generate suggestions for improvements and new services.

[0928] Input: Customer feedback, behavioral data

[0929] How it works: The server analyzes negative customer feedback and uses machine learning algorithms to identify areas for product improvement. It also automatically generates new service proposals.

[0930] Output: List of improvement proposals, new service proposals

[0931] The above are the specific processing steps of this system.

[0932] (Application example 1)

[0933] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0934] In recent years, companies have been able to obtain vast amounts of data, but there is a demand for efficient systems that can effectively utilize this data and help with business decision-making. In particular, brick-and-mortar stores lack the means to quickly analyze data collected in real time and develop sales forecasts and marketing strategies. Furthermore, there is a demand for rapid response by providing store staff with analysis of customer feedback and demand forecasts in real time. To address these challenges, the present invention provides advanced data processing technology and a means for visualizing data in real time.

[0935] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0936] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and shaping the data, means for analyzing the shaped data using natural language processing technology and extracting important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction, means for cleaning the collected data in real time and performing analysis, sentiment analysis, and topic modeling using natural language processing technology, means for making demand forecasts based on past sales data and providing sales forecasts for the next period, and means for displaying data visualizations on smart glasses and notifying important trends and alerts in real time. This enables companies to extract useful information from massive amounts of data and use it for customer analysis, demand forecasting, and sales management.

[0937] A "data source" is the source of the original data used to collect information, and includes social media sites and customer review sites.

[0938] "Means for collecting information" refers to the techniques and methods for efficiently acquiring the necessary information from data sources.

[0939] A "database" is a collection of structured data that is used to systematically store and manage collected information.

[0940] "Information cleaning methods" are techniques and methods for correcting missing data and outliers from collected data and improving the quality of the data.

[0941] "Means for data preparation" refers to techniques and methods for converting cleaned data into an analyzable form and formatting it appropriately.

[0942] "Natural language processing technology" is a computer technology for processing human language, including text tokenization, sentiment analysis, and topic modeling.

[0943] "Means for extracting important keywords and patterns" refers to techniques and methods for finding important information and trends that are useful for business from analyzed data.

[0944] A "means for training a predictive model" is a technique or method for building and training a model to predict future events or conditions based on past data.

[0945] "Means for making future predictions" are technologies or methods that use trained predictive models to forecast future demand, sales, etc.

[0946] "Means for visualizing in dashboard format" refers to techniques and methods for displaying analysis results and prediction results in a format that is intuitively easy to understand, including graphs and charts.

[0947] "Means for displaying on a device" refers to the technology or method for displaying the information visualized in dashboard format on a device used by the user (e.g., smart glasses, tablet).

[0948] "Means for analyzing customer feedback and behavioral data" refers to techniques and methods for analyzing customer opinions and behavioral history to evaluate customer satisfaction and behavioral trends.

[0949] "Means for proposing measures to improve customer satisfaction" refers to techniques and methods that propose specific actions to improve customer satisfaction based on customer feedback and behavioral data.

[0950] "Means for cleaning data collected in real time" refers to techniques and methods for immediately correcting missing or outliers in data collected in real time, thereby maintaining the quality of the data.

[0951] "Means for sentiment analysis and topic modeling" refers to techniques and methods that use natural language processing techniques to analyze sentiment trends and major topics from text.

[0952] "Means for forecasting demand based on past sales data" refers to techniques and methods for forecasting future sales and demand using historical sales data.

[0953] The "means for providing a sales forecast for the next period" refers to a technique or method for predicting future sales for a specific period and providing the results.

[0954] "Means for displaying on smart glasses" means the technology or method for displaying visualized data or alerts on smart glasses, allowing users to view the information in real time.

[0955] "Means for notifying important trends and alerts" refers to technologies and methods that immediately notify users of trends or anomalies that require particular attention based on the results of data analysis.

[0956] The "data analysis system for brick-and-mortar stores" of the present invention effectively collects huge amounts of data and analyzes it in real time to optimize brick-and-mortar store operations. Specific embodiments of this system are described below.

[0957] 1. Data Collection

[0958] The server collects customer feedback and sales data from social media, customer review sites, internal CRM systems, etc. For example, it uses the Twitter API to retrieve tweets related to product names and collect customer opinions.

[0959] 2. Data cleaning and formatting

[0960] The collected data is cleaned using the Pandas library to correct missing data and outliers and improve data quality. This cleaned data is then formatted using natural language processing techniques, specifically NLTK and SpaCy, to tokenize the text and convert it into a parseable format.

[0961] 3. Data analysis

[0962] The server uses natural language processing technology to perform sentiment analysis and topic modeling on the formatted data. For example, it analyzes customer tweets and determines whether their content is positive or negative. This allows for a quantitative understanding of product reviews on social media.

[0963] 4. Demand forecasting

[0964] Based on past sales data, the server uses Scikit-learn to train a predictive model. For example, monthly sales data is used to forecast sales for the next three months. The trained model can then predict future demand with high accuracy.

[0965] 5. Data Visualization and Notification

[0966] The server visualizes the prediction and analysis results in the form of a dashboard. Graphs and charts are created using Matplotlib and Seaborn. The generated visualization data is displayed in real time on smart glasses via AR Kit or the Google Glass API. This allows users to receive the information they need to run their stores in real time and instantly grasp important trends and alerts.

[0967] Specific examples

[0968] As a concrete example, consider a store planning to launch a new product. In this system, the server collects information about the new product from social media and customer review sites and performs sentiment analysis. For example, if a prompt such as "Collect social media reactions about product X and display the results of the sentiment analysis on the smart glasses in real time" is input into the generative AI model, the results of the sentiment analysis will be displayed in real time on the smart glasses. In addition, sales forecasts for the next three months are generated based on past sales data and notified to the user. This can be used to develop marketing strategies and production plans for the new product.

[0969] The present invention is a system that eliminates the difficulty of understanding information due to the vast amount of data and provides the ability to make evidence-based decisions, thereby significantly improving the competitiveness of physical stores.

[0970] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0971] Processing Steps

[0972] Step 1:

[0973] The server collects information from data sources (social media, customer review sites, internal CRM systems, etc.).

[0974] Input: API key or authentication information for data collection, search keywords.

[0975] Output: The raw data collected.

[0976] Specific operation: The server uses the Twitter API to search for tweets related to "product name" and obtains the results in text format.

[0977] Step 2:

[0978] The server stores the collected raw data in a database.

[0979] Input: Collected raw data (text data).

[0980] Output: The saved database entries.

[0981] Specific operation: The server converts the tweet data into a tabular format using a Pandas data frame and saves it in an SQL database.

[0982] Step 3:

[0983] The server cleans the stored data and corrects missing or outlier values.

[0984] Input: A saved database entry.

[0985] Output: Cleaned data.

[0986] Specific operation: The server uses the Pandas library to remove or impute missing data and correct outliers.

[0987] Step 4:

[0988] The server formats the cleaned data and applies natural language processing techniques.

[0989] Input: Cleaned data.

[0990] Output: Formatted text data.

[0991] Specific operation: The server uses the NLTK or SpaCy library to tokenize the text data, remove stop words, and extract the necessary features.

[0992] Step 5:

[0993] The server analyzes the data using natural language processing techniques to perform sentiment analysis and topic modeling.

[0994] Input: Formatted text data.

[0995] Output: Analysis results (sentiment scores, topic distribution).

[0996] Specific operation: The server performs sentiment analysis of the text using NLTK's SentimentIntensityAnalyzer and extracts important topics using an LDA model.

[0997] Step 6:

[0998] The server trains a predictive model based on past sales data and makes future predictions.

[0999] Input: Historical sales data, cleaned data.

[1000] Output: Sales forecast results.

[1001] What happens: The server uses Scikit-learn to train a linear regression model and perform sales forecasts for the next three months.

[1002] Step 7:

[1003] The server visualizes the prediction and analysis results in dashboard format and displays them on the terminal.

[1004] Input: Sales forecast results, sentiment analysis and topic modeling results.

[1005] Output: Visualized graphs on a dashboard.

[1006] Specific operation: The server creates graphs and charts using Matplotlib and Seaborn and generates a web dashboard.

[1007] Step 8:

[1008] The server analyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction.

[1009] Input: Customer feedback data, behavioral data.

[1010] Output: Specific action proposals to improve customer satisfaction.

[1011] Specific operation: The server performs sentiment analysis based on the feedback data, extracts areas for improvement in products and services, and notifies the marketing team.

[1012] Step 9:

[1013] The server cleans the collected data in real time and uses natural language processing technology to perform analysis, sentiment analysis and topic modeling.

[1014] Input: Raw data collected in real time.

[1015] Output: Cleaned and analyzed data.

[1016] Specific operation: The server cleans real-time streaming data with Pandas, analyzes sentiment scores with NLTK, and extracts topics with an LDA model.

[1017] Step 10:

[1018] The server makes a demand forecast based on past sales data and provides a sales forecast for the next period.

[1019] Input: Historical sales data.

[1020] Output: Sales forecast for next period.

[1021] What it does: The server uses Scikit-learn's time series forecasting model to forecast future sales and generate results.

[1022] Step 11:

[1023] The server displays data visualizations on the smart glasses, notifying users of important trends and alerts in real time.

[1024] Inputs: Visualization data, important trends or alerts.

[1025] Output: Real-time data displayed on smart glasses.

[1026] What it does: The server uses AR Kit or Google Glass APIs to display important data on the smart glasses and notify you of trends and alerts.

[1027] Through the above steps, this system can effectively collect and analyze huge amounts of data, significantly supporting decision-making in brick-and-mortar store operations.

[1028] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1029] The present invention is a system that aims to enable companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, making it possible to present more sophisticated measures to improve customer satisfaction. The program processing of this system is explained in detail below.

[1030] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[1031] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[1032] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[1033] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[1034] The server visualizes the prediction and analysis results in the form of a dashboard and displays it on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[1035] Furthermore, in this invention, the server analyzes customer feedback and behavioral data to present measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user. Here, the emotion engine plays a particularly important role. The emotion engine recognizes the user's emotions and stores the emotion data in a database. The server also analyzes this emotion data to assist in providing more precise measures to improve customer satisfaction.

[1036] Specific examples

[1037] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[1038] The server uses NLP techniques to analyze customer feedback, perform sentiment analysis and topic modeling, analyze behavioral data, and perform clustering to identify customer buying patterns, which allows the server to identify desired features and shortcomings of new products.

[1039] The server then trains a predictive model, forecasting demand for new products based on past data and estimating sales figures for the next three months, for example. The forecast and analysis results are then visualized in a dashboard format and displayed on the device.

[1040] Users can check the results through the dashboard and develop specific marketing strategies and production plans, while also referring to the data from the emotion engine provided by the server. The emotion engine analyzes customer emotions, and the server uses this data to suggest specific improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[1041] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information caused by vast amounts of data, make evidence-based decisions, and improve efficiency and competitiveness.In addition, by utilizing the emotion engine, it can further contribute to improving customer satisfaction.

[1042] The processing flow will be explained below.

[1043] Step 1:

[1044] The server collects information from data sources, such as social media, customer review sites, and internal CRM systems. The server uses data collection APIs to retrieve data from each source and stores it in a database.

[1045] Step 2:

[1046] The server cleans the stored data, removing missing data and noise, and detecting and correcting outliers, for example by imputing missing values ​​with the mean or correcting outliers to improve the quality of the data.

[1047] Step 3:

[1048] The server formats the data, converts it into a format that is easy to analyze, extracts the necessary features, and normalizes the data so that all data is handled on a uniform scale.

[1049] Step 4:

[1050] The server uses natural language processing (NLP) to analyze the text data, tokenize customer feedback, extract key keywords and phrases, and perform sentiment analysis and topic modeling to understand trends in customer opinions and sentiment.

[1051] Step 5:

[1052] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify patterns in user behavior. It uses a clustering algorithm to group users with similar behavior.

[1053] Step 6:

[1054] The server uses an emotion engine to recognize the user's emotions. For example, it analyzes emotions from customer text feedback and facial expression data. The emotion data is stored in a database and integrated with other data for analysis.

[1055] Step 7:

[1056] The server trains the predictive model. Based on past data, it uses time series analysis and regression analysis models to predict future market demand and customer behavior. The predictive model is then evaluated for accuracy through training and testing on data.

[1057] Step 8:

[1058] The server performs the prediction execution, using the trained model to make specific predictions about future events, such as estimating sales figures or customer intent after a new product launch.

[1059] Step 9:

[1060] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts and displays the information on an interactive dashboard.

[1061] Step 10:

[1062] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[1063] Step 11:

[1064] The server proposes measures to improve customer satisfaction based on data from the emotion engine. It comprehensively analyzes customer feedback, behavioral data, and recognized emotional data to propose specific improvements to products and services. The proposals are presented via the device, and the user can then take action to increase customer satisfaction.

[1065] Example 2

[1066] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1067] Modern companies face challenges in efficiently collecting massive amounts of data, analyzing it, and making future predictions and decisions. Furthermore, a lack of precise ways to recognize customer sentiment and improve customer satisfaction makes it difficult to properly understand customer needs and respond quickly. Existing systems perform data cleaning, natural language processing analysis, and predictive model training separately, making integrated analysis and prediction difficult, and making effective decision-making difficult.

[1068] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1069] In this invention, the server includes: [means for collecting information from data sources;] [means for storing the collected information in a database;] [means for cleaning the stored information and formatting the data;] [means for analyzing the formatted data using natural language processing technology and extracting important keywords and patterns;] [means for training a predictive model based on past data and making future predictions;] [means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal; and [means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction using an emotion engine.] This enables companies to collect and analyze data in an integrated manner, make future predictions and decisions, and quickly and effectively respond to customer needs and improve customer satisfaction.

[1070] "Data sources" are sources of data collected from external and internal systems and platforms, such as social media, customer review sites, and CRM systems.

[1071] A "database" is an information system that stores collected information and enables it to be efficiently managed and searched. Examples include relational databases and NoSQL databases.

[1072] "Cleaning" is the process of correcting and removing missing data and outliers from stored information to improve data quality.

[1073] "Natural language processing technology" refers to computer techniques for analyzing, understanding, and generating human language. Specifically, it includes tokenization, sentiment analysis, and topic modeling.

[1074] A "predictive model" is a statistical or machine learning model trained to predict future trends based on past data. It includes techniques such as time series analysis and regression analysis.

[1075] A "dashboard" is an interface that visually displays analysis and prediction results, allowing users to intuitively grasp the situation.

[1076] An "emotion engine" is a technology or algorithm that analyzes user emotions and generates corresponding emotional data. It helps to precisely improve customer satisfaction.

[1077] "Customer satisfaction improvement measures" are specific action plans for presenting improvements and new proposals for products and services based on customer evaluations and feedback.

[1078] The present invention is a system that enables companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, by incorporating an emotion engine that recognizes user emotions, it is possible to provide more sophisticated measures to improve customer satisfaction. Specific embodiments of this system are described below.

[1079] First, the server collects information from data sources. Data sources include social media, customer review sites, and internal CRM systems. For example, the server might use an API to collect data related to a specific hashtag from a social media site. The server then converts the collected information into an appropriate format and stores it in a database. Relational databases or NoSQL databases are often used for this purpose.

[1080] As a concrete example, a server uses the Twitter API to collect tweets containing the hashtag "new product release," converts them into JSON format, and stores them in a database.

[1081] Next, the server cleans the stored information. This step uses Python's Pandas library to impute missing data and correct outliers. For example, records with missing purchase dates are imputed with the user's closest purchase date. Additionally, any outliers that exist are removed through manual review or automated correction.

[1082] The server then formats the data, selects and normalizes the necessary features, for example, converting customer age, purchase frequency, and purchase amount to a standardized scale using scaling and feature selection techniques. The formatted data is then split into a training dataset and a test dataset.

[1083] The server then uses natural language processing (NLP) techniques to analyze this data. Specifically, it uses the SpaCy library to tokenize customer feedback and extract key keywords. It also performs sentiment analysis to assign a positive, negative, or neutral sentiment score to the feedback. Finally, it uses Scikit-learn to cluster behavioral data and identify customer purchasing patterns.

[1084] Next, the server uses TensorFlow to train a predictive model for time series analysis. For example, it predicts sales figures for the next three months based on past data. The predictive model sets evaluation indicators and evaluates reliability. If anomalies are detected, the model can be retrained.

[1085] The server uses Tableau to visualize these predictions and analysis results in dashboard format and displays them on the device. Users can access the dashboard using a PC or tablet and intuitively grasp the situation, which is useful for formulating marketing strategies and production plans.

[1086] Furthermore, the server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The emotion engine analyzes the user's emotions and stores the results in a database. The server then generates measures to improve customer satisfaction based on the emotion data and proposes them to the user.

[1087] Examples of prompt statements

[1088] How do I impute missing data and remove outliers using Python's Pandas?

[1089] "Describe an effective method for conducting sentiment analysis from customer feedback using natural language processing."

[1090] "How do I train a time series analysis model using TensorFlow to forecast sales for the next three months?"

[1091] "Please explain how to visualize data collected with Tableau in the form of a dashboard."

[1092] This allows companies to overcome the difficulty of comprehending information caused by the sheer volume of data, make evidence-based decisions, and quickly implement concrete actions to improve customer satisfaction.

[1093] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1094] Step 1:

[1095] The server collects information from data sources, including social media, customer review sites, and internal CRM systems. It retrieves data as input by sending API requests. Specifically, it uses the Twitter API to collect tweets containing the hashtag "new product release" and converts the retrieved data into JSON format. It generates a JSON object of the collected data as output.

[1096] Step 2:

[1097] The server stores the collected information in a database. The input is the JSON object collected in step 1. Specifically, it executes an INSERT query against an SQL or NoSQL database to store the data. The output is the number of records stored in the database.

[1098] Step 3:

[1099] The server cleans the stored information. The input is raw data read from the database. Specifically, it uses Python's Pandas library to impute missing data and correct outliers. For example, it imputes missing purchase dates with the user's closest purchase date. The output is a cleaned data frame.

[1100] Step 4:

[1101] The server formats the cleaned data. The input is a cleaned data frame. Specifically, features such as customer age, purchase frequency, and purchase amount are selected and normalized. Feature scaling is performed using Scikit-learn's StandardScaler. The output is a formatted data frame.

[1102] Step 5:

[1103] The server uses natural language processing techniques to analyze the formatted data. The input is a formatted data frame. Specifically, the SpaCy library is used to tokenize the customer feedback and extract important keywords. A sentiment analysis engine is used to calculate a sentiment score for the feedback. The output is a new data frame containing the analysis results.

[1104] Step 6:

[1105] The server clusters the behavioral data. The input is a data frame analyzed using natural language processing techniques. Specifically, it uses Scikit-learn to perform K-means clustering to identify customer purchasing patterns. The output is the data points corresponding to each cluster and the coordinates of the cluster center.

[1106] Step 7:

[1107] The server trains a predictive model. The input is clustered historical data. TensorFlow is used to train a predictive model for time series analysis. For example, a model is created to predict sales for the next three months. The output is a dataset containing the predicted values.

[1108] Step 8:

[1109] The server visualizes the prediction and analysis results in the form of a dashboard and presents it to the terminal. The input is a dataset containing the prediction values ​​and analysis results. Specifically, the dashboard is created using Tableau and provided to the end user through a web-based interface. The output is a visualized dashboard that can be viewed on the terminal.

[1110] Step 9:

[1111] The server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The input is customer feedback data. Specifically, the emotion engine analyzes user feedback and generates an emotion score. This is then stored in a database, and measures to improve customer satisfaction are generated based on the analysis results. The output is a report proposing measures to improve customer satisfaction.

[1112] Through these steps, the system enables businesses to efficiently collect and analyze data and provide actionable insights, which in turn contributes to improving customer satisfaction.

[1113] (Application example 2)

[1114] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1115] In recent years, companies have been required to quickly and accurately collect and analyze massive amounts of data and make appropriate decisions based on that data. However, due to the sheer volume and diversity of data, there is a lack of concrete methods for how to collect, analyze, and utilize this information. Furthermore, while there is a need to further improve customer satisfaction by recognizing customer emotions in real time, current systems have difficulty doing this efficiently. This creates the challenge of being unable to grasp the detailed needs and emotions of customers, making it difficult to present appropriate product proposals and improvement measures.

[1116] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1117] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and formatting the data, means for analyzing the formatted data using natural language processing technology to extract important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, means for analyzing customer reviews and emotion data in real time and recommending products based on related topics and emotion scores, and means for visualizing emotion score distributions and clustering results on the dashboard. This enables companies to efficiently collect and analyze massive amounts of data, recognize customer emotions in real time, and propose product suggestions and improvement measures.

[1118] A "data source" is a data provider that includes a file, database, or other source of information.

[1119] "Means for collecting information" refers to the tools and techniques used to obtain the necessary data from designated data sources and input it into the system.

[1120] "Means for storing information in a database" refers to the technology or system for storing collected data in an appropriate format as structured data.

[1121] "Information cleaning and data formatting methods" are processes and techniques that correct incomplete or erroneous data and convert it into a coherent form.

[1122] "Natural language processing technology" is a technology that allows computers to understand, process, and analyze human language.

[1123] A "predictive model" is a mathematical or statistical model used to predict future behavior or outcomes based on past data.

[1124] A "dashboard format" is an interface or layout that displays data in a visually easy-to-understand way.

[1125] "Customer feedback" refers to customer evaluations and opinions about products and services.

[1126] "Customer behavioral data" refers to data related to customer behavior, such as purchasing history and website browsing history.

[1127] "Customer reviews" are ratings and comments written by customers about products or services.

[1128] "Emotional data" is emotional information extracted from customer reviews and feedback.

[1129] "Real-time analysis" is the process of analyzing data immediately at the moment it is generated.

[1130] "Relevant topics" are the main discussion themes or topics that emerge from the analyzed data.

[1131] An "emotion score" is a numerical representation of the degree of positivity or negativity of emotions extracted from text data.

[1132] "Clustering" is a technique for grouping data based on similarity.

[1133] The present invention is a system for businesses to efficiently collect and analyze data and improve customer satisfaction. The system incorporates data collection, cleaning, natural language processing, predictive model training, visualization with dashboards, sentiment analysis, and clustering functions.

[1134] System Configuration

[1135] 1. Data Collection and Storage

[1136] The server collects data from multiple data sources (e.g., social media, customer review sites, CRM systems, etc.) and stores it in a database. This data collection is done using the requests library.

[1137] 2. Data cleaning and formatting

[1138] The server cleans and formats the collected data. This step involves correcting missing data and outliers to improve data quality. This is done using the Python standard library and Pandas.

[1139] 3. Natural Language Processing and Sentiment Analysis

[1140] The server analyzes the cleaned data using natural language processing (NLP) techniques. Specifically, it performs sentiment analysis using TextBlob to extract important keywords and patterns. It also uses KMeans from scikit-learn for clustering.

[1141] 4. Training a predictive model and forecasting the future

[1142] The server uses historical data to train predictive models that forecast future market demand and customer behavior, using algorithms such as time series analysis and regression analysis.

[1143] 5. Visualizing and Presenting Results

[1144] The prediction and analysis results are visualized in a dashboard format. Users can access the dashboard and check the results using a PC or tablet. Visualization is performed using libraries such as Matplotlib and Dash.

[1145] 6. Proposing measures to improve customer satisfaction

[1146] The server analyzes customer feedback and behavioral data to suggest satisfaction improvement measures tailored to each individual customer. The emotion engine analyzes customer reviews and emotion data in real time and recommends products based on related topics and emotion scores.

[1147] Specific examples

[1148] For example, when a user of the "Online Shopping Improvement Assistant" on the online shopping site views a specific product, such as a "Bluetooth speaker," the server analyzes the reviews on that page. If there are many negative sentiments based on sentiment analysis, the server automatically notifies the user about areas that need improvement. Related topics (e.g., sound quality, battery life, connectivity) extracted from the reviews through clustering are also presented.

[1149] Prompt Sentence Examples

[1150] "Write a Python program that extracts negative sentiment from Bluetooth speaker reviews and notifies users with related topics. Use the TextBlob and scikit-learn libraries."

[1151] These systems enable companies to effectively utilize vast amounts of data and provide specific product suggestions and improvement measures based on customer sentiment, while users can check the analysis results in real time and make decisions quickly.

[1152] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1153] Program processing steps

[1154] Step 1:

[1155] The server collects information from data sources. Specifically, it collects data from multiple data sources (such as social media, customer review sites, and internal CRM systems) through APIs. It primarily uses the requests library to retrieve data in JSON format. The input is the URL and API key of the data source, and the output is the collected raw data.

[1156] Step 2:

[1157] Store the collected information in a database. The server converts the raw data into the required format and stores it in an appropriate database (e.g., MySQL or MongoDB). The input is the collected raw data, and the output is structured data stored in the database.

[1158] Step 3:

[1159] The stored information is cleaned and the data is formatted. The server uses the Pandas library to correct incomplete data and outliers and standardize the data format. The input is raw data retrieved from the database, and the output is cleaned, high-quality data.

[1160] Step 4:

[1161] The formatted data is analyzed using natural language processing techniques. The server uses the TextBlob library to analyze the text data of each review and calculate a sentiment score. It also performs clustering using scikit-learn. The input is the cleaned data, and the output is the sentiment score and clustering results for each review.

[1162] Step 5:

[1163] The server trains a predictive model based on historical data to predict market demand and customer behavior using a predictive algorithm. The input is the training data, and the output is the trained predictive model.

[1164] Step 6:

[1165] The prediction and analysis results are visualized in dashboard format and displayed on the terminal. The server uses libraries such as Matplotlib and Dash to visualize the sentiment scores, clustering results, and prediction results in graphs and charts. The input is the prediction model and analysis data, and the output is a visually easy-to-understand dashboard.

[1166] Step 7:

[1167] Customer feedback and behavioral data are analyzed to suggest measures to improve customer satisfaction. The server generates a specific action plan to improve customer satisfaction based on the analysis results and presents it to the user. The input is feedback data and behavioral data, and the output is a detailed report on areas for improvement and new proposals.

[1168] Step 8:

[1169] The system analyzes customer reviews and sentiment data in real time and makes product recommendations based on related topics and sentiment scores. The server extracts sentiment data from reviews and analyzes it in real time. The input is the review text data, and the output is product recommendations and related topics based on sentiment scores.

[1170] Step 9:

[1171] The emotion score distribution and clustering results are visualized on a dashboard. The server updates the dashboard so that users can check the emotion scores and clustering results in real time. The inputs are the emotion scores and clustering results, and the output is the visualized data on the dashboard.

[1172] By combining these processing steps, companies can efficiently collect and analyze huge amounts of data and take concrete measures to improve customer satisfaction.

[1173] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1174] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1175] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1176] [Fourth embodiment]

[1177] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1178] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1179] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1180] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1181] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1182] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1183] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1184] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1185] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1186] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1187] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1188] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1189] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1190] The present invention is a system that enables companies to efficiently process huge amounts of data and support future predictions and decision-making. The program processing of this system will be specifically described below.

[1191] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[1192] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[1193] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[1194] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[1195] The server visualizes the prediction and analysis results in a dashboard format and displays them on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[1196] Furthermore, the server analyzes customer feedback and behavioral data to suggest measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user, allowing the user to take appropriate actions to increase customer satisfaction.

[1197] Specific examples

[1198] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[1199] The server then analyzes customer feedback using NLP techniques, performing sentiment analysis and topic modeling, analyzing behavioral data and conducting clustering to identify customer purchasing patterns, which allows the server to identify desired features and shortcomings of new products.

[1200] The server trains a predictive model based on past data to forecast demand for new products. For example, it can estimate sales for the next three months. The forecast and analysis results are visualized in a dashboard format and displayed on the device.

[1201] Users can check the results through a dashboard and develop specific marketing strategies and production plans, taking into account the insights provided by the server. Furthermore, the server analyzes customer feedback and provides specific suggestions for improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[1202] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information due to the vast amount of data, make evidence-based decisions, and improve efficiency and competitiveness. It also contributes to improving customer satisfaction.

[1203] The processing flow will be explained below.

[1204] Step 1:

[1205] The server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server uses a data collection API to retrieve data from each source and stores it in a database.

[1206] Step 2:

[1207] The server cleans the data. The collected data may contain missing values ​​or outliers. The server detects these data and either fills in the missing values ​​with the average value or corrects the outliers. The server checks the data consistency to improve the data quality.

[1208] Step 3:

[1209] The server formats the data. To format the data in a form that is easy to analyze, it selects features and normalizes the data. In this step, the server extracts the necessary columns and standardizes the scale of the numerical data.

[1210] Step 4:

[1211] The server uses natural language processing technology to analyze the text data. For example, it tokenizes customer feedback and extracts important keywords. It also performs sentiment analysis and topic modeling to understand trends in customer opinions and emotions.

[1212] Step 5:

[1213] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify user segments and purchasing patterns. It uses clustering algorithms to identify groups of users with similar behavior.

[1214] Step 6:

[1215] The server trains the predictive model. It trains time series analysis models and regression analysis models based on past data to predict future market demand and customer behavior. In this step, data is trained and tested to evaluate the performance of the model.

[1216] Step 7:

[1217] The server performs the prediction execution, using the trained model to predict future events, such as estimating sales figures or customer intent after a new product launch.

[1218] Step 8:

[1219] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts to create an interactive dashboard.

[1220] Step 9:

[1221] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[1222] Step 10:

[1223] The server presents measures to improve customer satisfaction. It analyzes customer feedback and behavioral data and makes specific suggestions for improving products and services. Users can take appropriate actions based on the suggestions presented via their devices.

[1224] Example 1

[1225] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1226] Modern companies need to collect huge amounts of data from social media, customer review sites, and other sources, but it is difficult to efficiently organize and analyze that data and use it for future predictions and decision-making. Furthermore, there is a lack of technology to automatically generate specific proposals to improve customer satisfaction by utilizing customer feedback and behavioral data, so these issues must be resolved.

[1227] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1228] In this invention, the server includes: means for collecting information from data sources such as social media, customer review sites, and internal CRM systems; means for converting the collected information into JSON or CSV format and storing it in a database; means for cleaning the stored information using a library such as Pandas and formatting the data; means for analyzing the formatted data using natural language processing technology to extract important keywords, sentiment, and topics; means for the server to analyze behavioral data using techniques such as K-means clustering and identify user segments and purchasing patterns; means for training ARIMA models and regression analysis using past data to predict future market demand and customer behavior; means for visualizing the prediction and analysis results in dashboard format using Matplotlib or Plotly and displaying them on a terminal; and means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction. This enables efficient processing of massive amounts of data, supporting corporate decision-making, and improving customer satisfaction.

[1229] A "data source" is the original location or medium where information is stored, such as social media, customer review sites, or internal CRM systems.

[1230] "Gathering" refers to the process by which the server obtains the required information from the data sources.

[1231] "JSON" and "CSV" are standard data formats for storing collected data.

[1232] A "database" is a storage system that systematically organizes and stores information, allowing it to be accessed and managed efficiently.

[1233] "Cleaning" is the process of correcting or removing missing or outlier values ​​in data.

[1234] The "Pandas library" is a library for data processing and analysis in the Python language.

[1235] "Formatting" is the process of converting data into a form that is easier to analyze.

[1236] "Natural language processing technology" refers to the technology that enables computers to understand and process human language.

[1237] "Keywords" are important words and phrases extracted from the data.

[1238] "Sentiment analysis" is the process of identifying and classifying emotions and opinions contained within text data.

[1239] "Topic modeling" is a technology that automatically extracts topics from large amounts of text data.

[1240] "K-means clustering" is a machine learning algorithm for dividing data into several clusters (groups).

[1241] A "user segment" is a collection or group of users who share common characteristics.

[1242] "Purchasing patterns" are behaviors and tendencies that can be identified from a user's past purchasing history.

[1243] The "ARIMA model" is a statistical model used to analyze and predict time series data.

[1244] "Regression analysis" is a statistical method for filling in the relationships between variables.

[1245] "Forecasting" is the estimation of future events based on past data.

[1246] "Visualization" refers to the visual display of data and its analysis results using graphs and charts.

[1247] A "dashboard" is an interface for displaying multiple pieces of information on a single screen.

[1248] "Customer feedback" refers to opinions and evaluations from customers.

[1249] "Measures to improve customer satisfaction" are specific measures and proposals to increase customer satisfaction.

[1250] This invention is a system that enables companies to efficiently process large amounts of data and support future predictions and decision-making. This system is composed of a server, terminals, and users.

[1251] Server Processing

[1252] The server first collects information from data sources such as social media, customer review sites, and internal CRM systems. This information is collected using scripts and APIs written in Python. For example, to use the Twitter API to collect tweets containing a specific hashtag, the following method is used:

[1253] The server converts the collected data into JSON or CSV format and stores it in a database. This unifies the information collected from different data sources and makes subsequent processing easier. The server also uses the Pandas library to clean the data, correcting missing values ​​and outliers.

[1254] Data Shaping and Analysis

[1255] The server then formats the data and converts it into a format suitable for analysis. For example, it tokenizes the text data of customer reviews and removes stop words. Using this formatted data, the server analyzes customer feedback using natural language processing techniques. This processing involves keyword extraction using Scikit-learn's TfidfVectorizer and sentiment analysis using NLTK.

[1256] Clustering and Prediction

[1257] The server then uses algorithms such as K-means clustering to analyze behavioral data and identify user segments and purchasing patterns. It also uses historical data to train ARIMA models and regression analysis to predict future market demand and customer behavior. For example, an ARIMA model is used to forecast sales for the next three months based on past sales data.

[1258] Visualization and Presentation

[1259] The server visualizes the prediction and analysis results in dashboard format using Matplotlib or Plotly. This allows the data sent to the terminal to be displayed in a visually easy-to-understand format. Users can use this dashboard to check the analysis and prediction results and develop specific marketing strategies and production plans. The server also reanalyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction to users. This allows users to easily receive suggestions for product improvements and new services.

[1260] Specific examples

[1261] For example, consider a company planning to launch a new product. The server uses the Twitter API to collect tweets containing the hashtag "new product" and stores them in a database. The Pandas library is then used to impute missing values ​​in the collected data, tokenize it, and remove stop words. Next, Scikit-learn and NLTK are used to extract important keywords and perform sentiment analysis. K-means is used for clustering, and an ARIMA model is used to predict sales for the next three months. These results are visualized with Matplotlib and sent to the device.

[1262] Prompt Sentence Examples

[1263] "Write a program that uses the Twitter API to collect tweets containing a specific hashtag, cleans the data using Pandas, extracts important keywords using TfidfVectorizer, and then uses an ARIMA model to forecast sales for the next three months."

[1264] As a result, this system can efficiently process huge amounts of data, support corporate decision-making, and improve customer satisfaction.

[1265] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1266] Step 1:

[1267] The server collects data from social media, customer review sites, and internal CRM systems.

[1268] Input: Twitter API, customer review site API, data access method for internal CRM system

[1269] Specific operation: The server uses the Twitter API to collect tweets containing the hashtag "new product." Next, it collects product reviews from customer review sites via the API and retrieves customer purchase history data from the internal CRM system.

[1270] Output: Collected raw data (JSON and CSV formats)

[1271] Step 2:

[1272] The server stores the collected data in a database.

[1273] Input: Collected raw data (JSON or CSV format)

[1274] Specific operation: The server converts the collected data into JSON or CSV format and saves it in the corresponding table in the database. For example, Twitter data is saved in a table called "SNS_data".

[1275] Output: Data stored in a database in a uniform format

[1276] Step 3:

[1277] The server cleans the data.

[1278] Input: Data stored in a database

[1279] Specific operation: The server uses the Pandas library to detect, correct, or remove missing and outlier values. For example, in a customer review dataset, it imputes missing values ​​with the mean value and filters out outliers.

[1280] Output: Cleaned dataframe

[1281] Step 4:

[1282] The server formats the data and converts it into a format suitable for analysis.

[1283] Input: Cleaned dataframe

[1284] What it does: The server tokenizes the customer review text data and removes stop words. Specifically, it vectorizes the data using Scikit-learn's TfidfVectorizer.

[1285] Output: Tokenized text data with stop words removed

[1286] Step 5:

[1287] The server analyzes the data using natural language processing technology.

[1288] Input: Tokenized text data with stop words removed

[1289] How it works: The server extracts important keywords, performs sentiment analysis using NLTK, and extracts major topics using LDA topic modeling.

[1290] Output: Keyword list, sentiment score, topic list

[1291] Step 6:

[1292] The server clusters the behavioral data.

[1293] Input: Organized behavioral data

[1294] How it works: The server uses the K-means clustering algorithm to group users' purchasing patterns, for example, dividing users into multiple clusters based on purchase frequency or purchase amount.

[1295] Output: User's cluster information

[1296] Step 7:

[1297] The server uses historical data to train the predictive model.

[1298] Input: Existing sales data and customer behavior data

[1299] What it does: The server uses an ARIMA model to forecast sales for the next three months, and also trains the model to forecast market demand using regression analysis.

[1300] Output: Sales forecast results, demand forecast results

[1301] Step 8:

[1302] The server visualizes the prediction and analysis results.

[1303] Input: Sales forecast results, keyword list, sentiment score, user cluster information

[1304] Specific operation: The server uses Matplotlib or Plotly to visualize the prediction results as line graphs, pie charts, and bar graphs.

[1305] Output: Visualized graphs and charts

[1306] Step 9:

[1307] The terminal presents the visualized results to the user.

[1308] Input: Visualized graphs and charts

[1309] Specific operation: The terminal launches the dashboard and displays visualized data on the screen. Users can access the dashboard via a PC or tablet and check the analysis results.

[1310] Output: Analysis results displayed in a dashboard

[1311] Step 10:

[1312] The server then re-analyzes customer feedback and behavioral data to generate suggestions for improvements and new services.

[1313] Input: Customer feedback, behavioral data

[1314] How it works: The server analyzes negative customer feedback and uses machine learning algorithms to identify areas for product improvement. It also automatically generates new service proposals.

[1315] Output: List of improvement proposals, new service proposals

[1316] The above are the specific processing steps of this system.

[1317] (Application example 1)

[1318] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1319] In recent years, companies have been able to obtain vast amounts of data, but there is a demand for efficient systems that can effectively utilize this data and help with business decision-making. In particular, brick-and-mortar stores lack the means to quickly analyze data collected in real time and develop sales forecasts and marketing strategies. Furthermore, there is a demand for rapid response by providing store staff with analysis of customer feedback and demand forecasts in real time. To address these challenges, the present invention provides advanced data processing technology and a means for visualizing data in real time.

[1320] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1321] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and shaping the data, means for analyzing the shaped data using natural language processing technology and extracting important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction, means for cleaning the collected data in real time and performing analysis, sentiment analysis, and topic modeling using natural language processing technology, means for making demand forecasts based on past sales data and providing sales forecasts for the next period, and means for displaying data visualizations on smart glasses and notifying important trends and alerts in real time. This enables companies to extract useful information from massive amounts of data and use it for customer analysis, demand forecasting, and sales management.

[1322] A "data source" is the source of the original data used to collect information, and includes social media sites and customer review sites.

[1323] "Means for collecting information" refers to the techniques and methods for efficiently acquiring the necessary information from data sources.

[1324] A "database" is a collection of structured data that is used to systematically store and manage collected information.

[1325] "Information cleaning methods" are techniques and methods for correcting missing data and outliers from collected data and improving the quality of the data.

[1326] "Means for data preparation" refers to techniques and methods for converting cleaned data into an analyzable form and formatting it appropriately.

[1327] "Natural language processing technology" is a computer technology for processing human language, including text tokenization, sentiment analysis, and topic modeling.

[1328] "Means for extracting important keywords and patterns" refers to techniques and methods for finding important information and trends that are useful for business from analyzed data.

[1329] A "means for training a predictive model" is a technique or method for building and training a model to predict future events or conditions based on past data.

[1330] "Means for making future predictions" are technologies or methods that use trained predictive models to forecast future demand, sales, etc.

[1331] "Means for visualizing in dashboard format" refers to techniques and methods for displaying analysis results and prediction results in a format that is intuitively easy to understand, including graphs and charts.

[1332] "Means for displaying on a device" refers to the technology or method for displaying the information visualized in dashboard format on a device used by the user (e.g., smart glasses, tablet).

[1333] "Means for analyzing customer feedback and behavioral data" refers to techniques and methods for analyzing customer opinions and behavioral history to evaluate customer satisfaction and behavioral trends.

[1334] "Means for proposing measures to improve customer satisfaction" refers to techniques and methods that propose specific actions to improve customer satisfaction based on customer feedback and behavioral data.

[1335] "Means for cleaning data collected in real time" refers to techniques and methods for immediately correcting missing or outliers in data collected in real time, thereby maintaining the quality of the data.

[1336] "Means for sentiment analysis and topic modeling" refers to techniques and methods that use natural language processing techniques to analyze sentiment trends and major topics from text.

[1337] "Means for forecasting demand based on past sales data" refers to techniques and methods for forecasting future sales and demand using historical sales data.

[1338] The "means for providing a sales forecast for the next period" refers to a technique or method for predicting future sales for a specific period and providing the results.

[1339] "Means for displaying on smart glasses" means the technology or method for displaying visualized data or alerts on smart glasses, allowing users to view the information in real time.

[1340] "Means for notifying important trends and alerts" refers to technologies and methods that immediately notify users of trends or anomalies that require particular attention based on the results of data analysis.

[1341] The "data analysis system for brick-and-mortar stores" of the present invention effectively collects huge amounts of data and analyzes it in real time to optimize brick-and-mortar store operations. Specific embodiments of this system are described below.

[1342] 1. Data Collection

[1343] The server collects customer feedback and sales data from social media, customer review sites, internal CRM systems, etc. For example, it uses the Twitter API to retrieve tweets related to product names and collect customer opinions.

[1344] 2. Data cleaning and formatting

[1345] The collected data is cleaned using the Pandas library to correct missing data and outliers and improve data quality. This cleaned data is then formatted using natural language processing techniques, specifically NLTK and SpaCy, to tokenize the text and convert it into a parseable format.

[1346] 3. Data analysis

[1347] The server uses natural language processing technology to perform sentiment analysis and topic modeling on the formatted data. For example, it analyzes customer tweets and determines whether their content is positive or negative. This allows for a quantitative understanding of product reviews on social media.

[1348] 4. Demand forecasting

[1349] Based on past sales data, the server uses Scikit-learn to train a predictive model. For example, monthly sales data is used to forecast sales for the next three months. The trained model can then predict future demand with high accuracy.

[1350] 5. Data Visualization and Notification

[1351] The server visualizes the prediction and analysis results in the form of a dashboard. Graphs and charts are created using Matplotlib and Seaborn. The generated visualization data is displayed in real time on smart glasses via AR Kit or the Google Glass API. This allows users to receive the information they need to run their stores in real time and instantly grasp important trends and alerts.

[1352] Specific examples

[1353] As a concrete example, consider a store planning to launch a new product. In this system, the server collects information about the new product from social media and customer review sites and performs sentiment analysis. For example, if a prompt such as "Collect social media reactions about product X and display the results of the sentiment analysis on the smart glasses in real time" is input into the generative AI model, the results of the sentiment analysis will be displayed in real time on the smart glasses. In addition, sales forecasts for the next three months are generated based on past sales data and notified to the user. This can be used to develop marketing strategies and production plans for the new product.

[1354] The present invention is a system that eliminates the difficulty of understanding information due to the vast amount of data and provides the ability to make evidence-based decisions, thereby significantly improving the competitiveness of physical stores.

[1355] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1356] Processing Steps

[1357] Step 1:

[1358] The server collects information from data sources (social media, customer review sites, internal CRM systems, etc.).

[1359] Input: API key or authentication information for data collection, search keywords.

[1360] Output: The raw data collected.

[1361] Specific operation: The server uses the Twitter API to search for tweets related to "product name" and obtains the results in text format.

[1362] Step 2:

[1363] The server stores the collected raw data in a database.

[1364] Input: Collected raw data (text data).

[1365] Output: The saved database entries.

[1366] Specific operation: The server converts the tweet data into a tabular format using a Pandas data frame and saves it in an SQL database.

[1367] Step 3:

[1368] The server cleans the stored data and corrects missing or outlier values.

[1369] Input: A saved database entry.

[1370] Output: Cleaned data.

[1371] Specific operation: The server uses the Pandas library to remove or impute missing data and correct outliers.

[1372] Step 4:

[1373] The server formats the cleaned data and applies natural language processing techniques.

[1374] Input: Cleaned data.

[1375] Output: Formatted text data.

[1376] Specific operation: The server uses the NLTK or SpaCy library to tokenize the text data, remove stop words, and extract the necessary features.

[1377] Step 5:

[1378] The server analyzes the data using natural language processing techniques to perform sentiment analysis and topic modeling.

[1379] Input: Formatted text data.

[1380] Output: Analysis results (sentiment scores, topic distribution).

[1381] Specific operation: The server performs sentiment analysis of the text using NLTK's SentimentIntensityAnalyzer and extracts important topics using an LDA model.

[1382] Step 6:

[1383] The server trains a predictive model based on past sales data and makes future predictions.

[1384] Input: Historical sales data, cleaned data.

[1385] Output: Sales forecast results.

[1386] What happens: The server uses Scikit-learn to train a linear regression model and perform sales forecasts for the next three months.

[1387] Step 7:

[1388] The server visualizes the prediction and analysis results in dashboard format and displays them on the terminal.

[1389] Input: Sales forecast results, sentiment analysis and topic modeling results.

[1390] Output: Visualized graphs on a dashboard.

[1391] Specific operation: The server creates graphs and charts using Matplotlib and Seaborn and generates a web dashboard.

[1392] Step 8:

[1393] The server analyzes customer feedback and behavioral data and suggests measures to improve customer satisfaction.

[1394] Input: Customer feedback data, behavioral data.

[1395] Output: Specific action proposals to improve customer satisfaction.

[1396] Specific operation: The server performs sentiment analysis based on the feedback data, extracts areas for improvement in products and services, and notifies the marketing team.

[1397] Step 9:

[1398] The server cleans the collected data in real time and uses natural language processing technology to perform analysis, sentiment analysis and topic modeling.

[1399] Input: Raw data collected in real time.

[1400] Output: Cleaned and analyzed data.

[1401] Specific operation: The server cleans real-time streaming data with Pandas, analyzes sentiment scores with NLTK, and extracts topics with an LDA model.

[1402] Step 10:

[1403] The server makes a demand forecast based on past sales data and provides a sales forecast for the next period.

[1404] Input: Historical sales data.

[1405] Output: Sales forecast for next period.

[1406] What it does: The server uses Scikit-learn's time series forecasting model to forecast future sales and generate results.

[1407] Step 11:

[1408] The server displays data visualizations on the smart glasses, notifying users of important trends and alerts in real time.

[1409] Inputs: Visualization data, important trends or alerts.

[1410] Output: Real-time data displayed on smart glasses.

[1411] What it does: The server uses AR Kit or Google Glass APIs to display important data on the smart glasses and notify you of trends and alerts.

[1412] Through the above steps, this system can effectively collect and analyze huge amounts of data, significantly supporting decision-making in brick-and-mortar store operations.

[1413] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1414] The present invention is a system that aims to enable companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, this system incorporates an emotion engine that recognizes the user's emotions, making it possible to present more sophisticated measures to improve customer satisfaction. The program processing of this system is explained in detail below.

[1415] First, the server collects information from data sources, such as social media, customer review sites, internal CRM systems, etc. The server then converts the collected information into the appropriate format and stores it in a database.

[1416] Next, the server cleans the stored information. In this step, missing data and outliers are corrected to improve the quality of the data. Furthermore, the server cleans the data, selecting and normalizing the necessary features.

[1417] The server then analyzes the formatted data using natural language processing techniques, such as tokenizing customer feedback to extract key keywords, sentiment analysis, and topic modeling, as well as analyzing behavioral data and performing clustering to identify user segments and purchasing patterns.

[1418] The server then trains a predictive model based on past data. This predictive model uses algorithms such as time series analysis and regression analysis to forecast future market demand and customer behavior. The server evaluates the reliability of the forecast results, correcting any outliers and retraining the model as necessary.

[1419] The server visualizes the prediction and analysis results in the form of a dashboard and displays it on the terminal. Users can access the dashboard using a PC or tablet to check the analysis and prediction results. This dashboard provides insights that are useful for formulating marketing strategies, production plans, and making business decisions.

[1420] Furthermore, in this invention, the server analyzes customer feedback and behavioral data to present measures to improve customer satisfaction. Specifically, it automatically generates product improvements and new service proposals and presents them to the user. Here, the emotion engine plays a particularly important role. The emotion engine recognizes the user's emotions and stores the emotion data in a database. The server also analyzes this emotion data to assist in providing more precise measures to improve customer satisfaction.

[1421] Specific examples

[1422] For example, consider a company planning to launch a new product. A server collects information about the new product from social media and customer review sites, as well as customer reactions to competing products. The collected information is stored in a database. The server then cleans and formats the data for accurate analysis.

[1423] The server uses NLP techniques to analyze customer feedback, perform sentiment analysis and topic modeling, analyze behavioral data, and perform clustering to identify customer buying patterns, which allows the server to identify desired features and shortcomings of new products.

[1424] The server then trains a predictive model, forecasting demand for new products based on past data and estimating sales figures for the next three months, for example. The forecast and analysis results are then visualized in a dashboard format and displayed on the device.

[1425] Users can check the results through the dashboard and develop specific marketing strategies and production plans, while also referring to the data from the emotion engine provided by the server. The emotion engine analyzes customer emotions, and the server uses this data to suggest specific improvements to new products, allowing users to take appropriate actions to increase customer satisfaction.

[1426] The above is an embodiment of the present invention, and this system allows companies to eliminate the difficulty of understanding information caused by vast amounts of data, make evidence-based decisions, and improve efficiency and competitiveness.In addition, by utilizing the emotion engine, it can further contribute to improving customer satisfaction.

[1427] The processing flow will be explained below.

[1428] Step 1:

[1429] The server collects information from data sources, such as social media, customer review sites, and internal CRM systems. The server uses data collection APIs to retrieve data from each source and stores it in a database.

[1430] Step 2:

[1431] The server cleans the stored data, removing missing data and noise, and detecting and correcting outliers, for example by imputing missing values ​​with the mean or correcting outliers to improve the quality of the data.

[1432] Step 3:

[1433] The server formats the data, converts it into a format that is easy to analyze, extracts the necessary features, and normalizes the data so that all data is handled on a uniform scale.

[1434] Step 4:

[1435] The server uses natural language processing (NLP) to analyze the text data, tokenize customer feedback, extract key keywords and phrases, and perform sentiment analysis and topic modeling to understand trends in customer opinions and sentiment.

[1436] Step 5:

[1437] The server analyzes the behavioral data. It analyzes users' website visits and purchase histories to identify patterns in user behavior. It uses a clustering algorithm to group users with similar behavior.

[1438] Step 6:

[1439] The server uses an emotion engine to recognize the user's emotions. For example, it analyzes emotions from customer text feedback and facial expression data. The emotion data is stored in a database and integrated with other data for analysis.

[1440] Step 7:

[1441] The server trains the predictive model. Based on past data, it uses time series analysis and regression analysis models to predict future market demand and customer behavior. The predictive model is then evaluated for accuracy through training and testing on data.

[1442] Step 8:

[1443] The server performs the prediction execution, using the trained model to make specific predictions about future events, such as estimating sales figures or customer intent after a new product launch.

[1444] Step 9:

[1445] The server visualizes the prediction and analysis results in the form of a dashboard, allowing users to instantly grasp important information. The server generates graphs and charts and displays the information on an interactive dashboard.

[1446] Step 10:

[1447] The terminal displays the dashboard, and users can access the dashboard using a PC or tablet to check the forecast and analysis results, which can be used to formulate marketing strategies and production plans.

[1448] Step 11:

[1449] The server proposes measures to improve customer satisfaction based on data from the emotion engine. It comprehensively analyzes customer feedback, behavioral data, and recognized emotional data to propose specific improvements to products and services. The proposals are presented via the device, and the user can then take action to increase customer satisfaction.

[1450] Example 2

[1451] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1452] Modern companies face challenges in efficiently collecting massive amounts of data, analyzing it, and making future predictions and decisions. Furthermore, a lack of precise ways to recognize customer sentiment and improve customer satisfaction makes it difficult to properly understand customer needs and respond quickly. Existing systems perform data cleaning, natural language processing analysis, and predictive model training separately, making integrated analysis and prediction difficult, and making effective decision-making difficult.

[1453] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1454] In this invention, the server includes: [means for collecting information from data sources;] [means for storing the collected information in a database;] [means for cleaning the stored information and formatting the data;] [means for analyzing the formatted data using natural language processing technology and extracting important keywords and patterns;] [means for training a predictive model based on past data and making future predictions;] [means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal; and [means for analyzing customer feedback and behavioral data and presenting measures to improve customer satisfaction using an emotion engine.] This enables companies to collect and analyze data in an integrated manner, make future predictions and decisions, and quickly and effectively respond to customer needs and improve customer satisfaction.

[1455] "Data sources" are sources of data collected from external and internal systems and platforms, such as social media, customer review sites, and CRM systems.

[1456] A "database" is an information system that stores collected information and enables it to be efficiently managed and searched. Examples include relational databases and NoSQL databases.

[1457] "Cleaning" is the process of correcting and removing missing data and outliers from stored information to improve data quality.

[1458] "Natural language processing technology" refers to computer techniques for analyzing, understanding, and generating human language. Specifically, it includes tokenization, sentiment analysis, and topic modeling.

[1459] A "predictive model" is a statistical or machine learning model trained to predict future trends based on past data. It includes techniques such as time series analysis and regression analysis.

[1460] A "dashboard" is an interface that visually displays analysis and prediction results, allowing users to intuitively grasp the situation.

[1461] An "emotion engine" is a technology or algorithm that analyzes user emotions and generates corresponding emotional data. It helps to precisely improve customer satisfaction.

[1462] "Customer satisfaction improvement measures" are specific action plans for presenting improvements and new proposals for products and services based on customer evaluations and feedback.

[1463] The present invention is a system that enables companies to efficiently collect and analyze large amounts of data and support future predictions and decision-making. Furthermore, by incorporating an emotion engine that recognizes user emotions, it is possible to provide more sophisticated measures to improve customer satisfaction. Specific embodiments of this system are described below.

[1464] First, the server collects information from data sources. Data sources include social media, customer review sites, and internal CRM systems. For example, the server might use an API to collect data related to a specific hashtag from a social media site. The server then converts the collected information into an appropriate format and stores it in a database. Relational databases or NoSQL databases are often used for this purpose.

[1465] As a concrete example, a server uses the Twitter API to collect tweets containing the hashtag "new product release," converts them into JSON format, and stores them in a database.

[1466] Next, the server cleans the stored information. This step uses Python's Pandas library to impute missing data and correct outliers. For example, records with missing purchase dates are imputed with the user's closest purchase date. Additionally, any outliers that exist are removed through manual review or automated correction.

[1467] The server then formats the data, selects and normalizes the necessary features, for example, converting customer age, purchase frequency, and purchase amount to a standardized scale using scaling and feature selection techniques. The formatted data is then split into a training dataset and a test dataset.

[1468] The server then uses natural language processing (NLP) techniques to analyze this data. Specifically, it uses the SpaCy library to tokenize customer feedback and extract key keywords. It also performs sentiment analysis to assign a positive, negative, or neutral sentiment score to the feedback. Finally, it uses Scikit-learn to cluster behavioral data and identify customer purchasing patterns.

[1469] Next, the server uses TensorFlow to train a predictive model for time series analysis. For example, it predicts sales figures for the next three months based on past data. The predictive model sets evaluation indicators and evaluates reliability. If anomalies are detected, the model can be retrained.

[1470] The server uses Tableau to visualize these predictions and analysis results in dashboard format and displays them on the device. Users can access the dashboard using a PC or tablet and intuitively grasp the situation, which is useful for formulating marketing strategies and production plans.

[1471] Furthermore, the server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The emotion engine analyzes the user's emotions and stores the results in a database. The server then generates measures to improve customer satisfaction based on the emotion data and proposes them to the user.

[1472] Examples of prompt statements

[1473] How do I impute missing data and remove outliers using Python's Pandas?

[1474] "Describe an effective method for conducting sentiment analysis from customer feedback using natural language processing."

[1475] "How do I train a time series analysis model using TensorFlow to forecast sales for the next three months?"

[1476] "Please explain how to visualize data collected with Tableau in the form of a dashboard."

[1477] This allows companies to overcome the difficulty of comprehending information caused by the sheer volume of data, make evidence-based decisions, and quickly implement concrete actions to improve customer satisfaction.

[1478] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1479] Step 1:

[1480] The server collects information from data sources, including social media, customer review sites, and internal CRM systems. It retrieves data as input by sending API requests. Specifically, it uses the Twitter API to collect tweets containing the hashtag "new product release" and converts the retrieved data into JSON format. It generates a JSON object of the collected data as output.

[1481] Step 2:

[1482] The server stores the collected information in a database. The input is the JSON object collected in step 1. Specifically, it executes an INSERT query against an SQL or NoSQL database to store the data. The output is the number of records stored in the database.

[1483] Step 3:

[1484] The server cleans the stored information. The input is raw data read from the database. Specifically, it uses Python's Pandas library to impute missing data and correct outliers. For example, it imputes missing purchase dates with the user's closest purchase date. The output is a cleaned data frame.

[1485] Step 4:

[1486] The server formats the cleaned data. The input is a cleaned data frame. Specifically, features such as customer age, purchase frequency, and purchase amount are selected and normalized. Feature scaling is performed using Scikit-learn's StandardScaler. The output is a formatted data frame.

[1487] Step 5:

[1488] The server uses natural language processing techniques to analyze the formatted data. The input is a formatted data frame. Specifically, the SpaCy library is used to tokenize the customer feedback and extract important keywords. A sentiment analysis engine is used to calculate a sentiment score for the feedback. The output is a new data frame containing the analysis results.

[1489] Step 6:

[1490] The server clusters the behavioral data. The input is a data frame analyzed using natural language processing techniques. Specifically, it uses Scikit-learn to perform K-means clustering to identify customer purchasing patterns. The output is the data points corresponding to each cluster and the coordinates of the cluster center.

[1491] Step 7:

[1492] The server trains a predictive model. The input is clustered historical data. TensorFlow is used to train a predictive model for time series analysis. For example, a model is created to predict sales for the next three months. The output is a dataset containing the predicted values.

[1493] Step 8:

[1494] The server visualizes the prediction and analysis results in the form of a dashboard and presents it to the terminal. The input is a dataset containing the prediction values ​​and analysis results. Specifically, the dashboard is created using Tableau and provided to the end user through a web-based interface. The output is a visualized dashboard that can be viewed on the terminal.

[1495] Step 9:

[1496] The server analyzes customer feedback and behavioral data and uses an emotion engine to suggest measures to improve customer satisfaction. The input is customer feedback data. Specifically, the emotion engine analyzes user feedback and generates an emotion score. This is then stored in a database, and measures to improve customer satisfaction are generated based on the analysis results. The output is a report proposing measures to improve customer satisfaction.

[1497] Through these steps, the system enables businesses to efficiently collect and analyze data and provide actionable insights, which in turn contributes to improving customer satisfaction.

[1498] (Application example 2)

[1499] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1500] In recent years, companies have been required to quickly and accurately collect and analyze massive amounts of data and make appropriate decisions based on that data. However, due to the sheer volume and diversity of data, there is a lack of concrete methods for how to collect, analyze, and utilize this information. Furthermore, while there is a need to further improve customer satisfaction by recognizing customer emotions in real time, current systems have difficulty doing this efficiently. This creates the challenge of being unable to grasp the detailed needs and emotions of customers, making it difficult to present appropriate product proposals and improvement measures.

[1501] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1502] In this invention, the server includes means for collecting information from data sources, means for storing the collected information in a database, means for cleaning the stored information and formatting the data, means for analyzing the formatted data using natural language processing technology to extract important keywords and patterns, means for training a predictive model based on past data and making future predictions, means for visualizing the prediction results and analysis results in dashboard format and presenting them on a terminal, means for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, means for analyzing customer reviews and emotion data in real time and recommending products based on related topics and emotion scores, and means for visualizing emotion score distributions and clustering results on the dashboard. This enables companies to efficiently collect and analyze massive amounts of data, recognize customer emotions in real time, and propose product suggestions and improvement measures.

[1503] A "data source" is a data provider that includes a file, database, or other source of information.

[1504] "Means for collecting information" refers to the tools and techniques used to obtain the necessary data from designated data sources and input it into the system.

[1505] "Means for storing information in a database" refers to the technology or system for storing collected data in an appropriate format as structured data.

[1506] "Information cleaning and data formatting methods" are processes and techniques that correct incomplete or erroneous data and convert it into a coherent form.

[1507] "Natural language processing technology" is a technology that allows computers to understand, process, and analyze human language.

[1508] A "predictive model" is a mathematical or statistical model used to predict future behavior or outcomes based on past data.

[1509] A "dashboard format" is an interface or layout that displays data in a visually easy-to-understand way.

[1510] "Customer feedback" refers to customer evaluations and opinions about products and services.

[1511] "Customer behavioral data" refers to data related to customer behavior, such as purchasing history and website browsing history.

[1512] "Customer reviews" are ratings and comments written by customers about products or services.

[1513] "Emotional data" is emotional information extracted from customer reviews and feedback.

[1514] "Real-time analysis" is the process of analyzing data immediately at the moment it is generated.

[1515] "Relevant topics" are the main discussion themes or topics that emerge from the analyzed data.

[1516] An "emotion score" is a numerical representation of the degree of positivity or negativity of emotions extracted from text data.

[1517] "Clustering" is a technique for grouping data based on similarity.

[1518] The present invention is a system for businesses to efficiently collect and analyze data and improve customer satisfaction. The system incorporates data collection, cleaning, natural language processing, predictive model training, visualization with dashboards, sentiment analysis, and clustering functions.

[1519] System Configuration

[1520] 1. Data Collection and Storage

[1521] The server collects data from multiple data sources (e.g., social media, customer review sites, CRM systems, etc.) and stores it in a database. This data collection is done using the requests library.

[1522] 2. Data cleaning and formatting

[1523] The server cleans and formats the collected data. This step involves correcting missing data and outliers to improve data quality. This is done using the Python standard library and Pandas.

[1524] 3. Natural Language Processing and Sentiment Analysis

[1525] The server analyzes the cleaned data using natural language processing (NLP) techniques. Specifically, it performs sentiment analysis using TextBlob to extract important keywords and patterns. It also uses KMeans from scikit-learn for clustering.

[1526] 4. Training a predictive model and forecasting the future

[1527] The server uses historical data to train predictive models that forecast future market demand and customer behavior, using algorithms such as time series analysis and regression analysis.

[1528] 5. Visualizing and Presenting Results

[1529] The prediction and analysis results are visualized in a dashboard format. Users can access the dashboard and check the results using a PC or tablet. Visualization is performed using libraries such as Matplotlib and Dash.

[1530] 6. Proposing measures to improve customer satisfaction

[1531] The server analyzes customer feedback and behavioral data to suggest satisfaction improvement measures tailored to each individual customer. The emotion engine analyzes customer reviews and emotion data in real time and recommends products based on related topics and emotion scores.

[1532] Specific examples

[1533] For example, when a user of the "Online Shopping Improvement Assistant" on the online shopping site views a specific product, such as a "Bluetooth speaker," the server analyzes the reviews on that page. If there are many negative sentiments based on sentiment analysis, the server automatically notifies the user about areas that need improvement. Related topics (e.g., sound quality, battery life, connectivity) extracted from the reviews through clustering are also presented.

[1534] Prompt Sentence Examples

[1535] "Write a Python program that extracts negative sentiment from Bluetooth speaker reviews and notifies users with related topics. Use the TextBlob and scikit-learn libraries."

[1536] These systems enable companies to effectively utilize vast amounts of data and provide specific product suggestions and improvement measures based on customer sentiment, while users can check the analysis results in real time and make decisions quickly.

[1537] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1538] Program processing steps

[1539] Step 1:

[1540] The server collects information from data sources. Specifically, it collects data from multiple data sources (such as social media, customer review sites, and internal CRM systems) through APIs. It primarily uses the requests library to retrieve data in JSON format. The input is the URL and API key of the data source, and the output is the collected raw data.

[1541] Step 2:

[1542] Store the collected information in a database. The server converts the raw data into the required format and stores it in an appropriate database (e.g., MySQL or MongoDB). The input is the collected raw data, and the output is structured data stored in the database.

[1543] Step 3:

[1544] The stored information is cleaned and the data is formatted. The server uses the Pandas library to correct incomplete data and outliers and standardize the data format. The input is raw data retrieved from the database, and the output is cleaned, high-quality data.

[1545] Step 4:

[1546] The formatted data is analyzed using natural language processing techniques. The server uses the TextBlob library to analyze the text data of each review and calculate a sentiment score. It also performs clustering using scikit-learn. The input is the cleaned data, and the output is the sentiment score and clustering results for each review.

[1547] Step 5:

[1548] The server trains a predictive model based on historical data to predict market demand and customer behavior using a predictive algorithm. The input is the training data, and the output is the trained predictive model.

[1549] Step 6:

[1550] The prediction and analysis results are visualized in dashboard format and displayed on the terminal. The server uses libraries such as Matplotlib and Dash to visualize the sentiment scores, clustering results, and prediction results in graphs and charts. The input is the prediction model and analysis data, and the output is a visually easy-to-understand dashboard.

[1551] Step 7:

[1552] Customer feedback and behavioral data are analyzed to suggest measures to improve customer satisfaction. The server generates a specific action plan to improve customer satisfaction based on the analysis results and presents it to the user. The input is feedback data and behavioral data, and the output is a detailed report on areas for improvement and new proposals.

[1553] Step 8:

[1554] The system analyzes customer reviews and sentiment data in real time and makes product recommendations based on related topics and sentiment scores. The server extracts sentiment data from reviews and analyzes it in real time. The input is the review text data, and the output is product recommendations and related topics based on sentiment scores.

[1555] Step 9:

[1556] The emotion score distribution and clustering results are visualized on a dashboard. The server updates the dashboard so that users can check the emotion scores and clustering results in real time. The inputs are the emotion scores and clustering results, and the output is the visualized data on the dashboard.

[1557] By combining these processing steps, companies can efficiently collect and analyze huge amounts of data and take concrete measures to improve customer satisfaction.

[1558] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1559] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1560] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1561] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1562] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1563] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1564] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1565] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1566] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1567] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1568] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1569] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1570] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1571] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1572] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1573] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1574] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1575] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1576] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1577] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1578] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1579] The following is further disclosed regarding the above embodiment.

[1580] (Claim 1)

[1581] [Means for collecting information from data sources and

[1582] [Means for storing the collected information in a database; and

[1583] [Means for cleaning the stored information and formatting the data;

[1584] [Means for analyzing formatted data using natural language processing technology and extracting important keywords and patterns;

[1585] [Means for training a predictive model based on past data and making future predictions;

[1586] [Means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal;

[1587] [Methods for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, and

[1588] A system including:

[1589] (Claim 2)

[1590] The system of claim 1, wherein the server is configured to clean the information collected from the data sources and analyze it using natural language processing techniques.

[1591] (Claim 3)

[1592] [The system according to claim 1, wherein the server trains a predictive model using past data and visualizes and presents the prediction results in a dashboard format.]

[1593] "Example 1"

[1594] (Claim 1)

[1595] [Methods for collecting information from data sources such as social media, customer review sites, and internal CRM systems;

[1596] [Means for converting the collected information into JSON or CSV format and saving it in a database, and

[1597] [Means for cleaning the stored information using the Pandas library, etc., and formatting the data,

[1598] [Means for analyzing formatted data using natural language processing technology to extract important keywords, sentiments, and topics;

[1599] [Means for the server to analyze behavioral data using methods such as K-means clustering to identify user segments and purchasing patterns;

[1600] [Methods for training ARIMA models and regression analysis using historical data to predict future market demand and customer behavior;

[1601] [Means for visualizing the prediction and analysis results in dashboard format using Matplotlib or Plotly and presenting them on the terminal,

[1602] [Methods for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, and

[1603] A system including:

[1604] (Claim 2)

[1605] The system of claim 1, wherein the server performs data cleaning using the Pandas library and analyzes customer feedback using natural language processing techniques.

[1606] (Claim 3)

[1607] [The system according to claim 1, further comprising a means for the server to train the ARIMA model using past data and visualize and present the prediction results in a dashboard format using Matplotlib.]

[1608] "Application Example 1"

[1609] (Claim 1)

[1610] [Means for collecting information from data sources and

[1611] [Means for storing the collected information in a database; and

[1612] [Means for cleaning the stored information and formatting the data;

[1613] [Means for analyzing formatted data using natural language processing technology and extracting important keywords and patterns;

[1614] [Means for training a predictive model based on past data and making future predictions;

[1615] [Means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal;

[1616] [Methods for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, and

[1617] [Means for cleaning the data collected in real time and performing analysis, sentiment analysis and topic modeling using natural language processing techniques;

[1618] [Means for forecasting demand based on past sales data and providing sales forecasts for the next period;

[1619] [A means to display data visualizations on smart glasses and notify important trends and alerts in real time; and

[1620] A system including:

[1621] (Claim 2)

[1622] The system of claim 1, wherein the server is configured to clean the information collected from the data sources and analyze it using natural language processing techniques.

[1623] (Claim 3)

[1624] [The system according to claim 1, wherein the server trains a predictive model using past data and visualizes and presents the prediction results in a dashboard format.]

[1625] "Example 2: Combining Emotion Engines"

[1626] (Claim 1)

[1627] [Means for collecting information from data sources and

[1628] [Means for storing the collected information in a database; and

[1629] [Means for cleaning the stored information and formatting the data;

[1630] [Means for analyzing formatted data using natural language processing technology and extracting important keywords and patterns;

[1631] [Means for training a predictive model based on past data and making future predictions;

[1632] [Means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal;

[1633] [Means for analyzing customer feedback and behavioral data and using an emotion engine to suggest measures to improve customer satisfaction;

[1634] A system including:

[1635] (Claim 2)

[1636] The system of claim 1, wherein the server is configured to clean the information collected from the data sources and analyze it using natural language processing techniques.

[1637] (Claim 3)

[1638] [The system according to claim 1, wherein the server trains a predictive model using past data and visualizes and presents the prediction results in a dashboard format.]

[1639] "Application example 2 when combining emotion engines"

[1640] (Claim 1)

[1641] [Means for collecting information from data sources and

[1642] [Means for storing the collected information in a database; and

[1643] [Means for cleaning the stored information and formatting the data;

[1644] [Means for analyzing formatted data using natural language processing technology and extracting important keywords and patterns;

[1645] [Means for training a predictive model based on past data and making future predictions;

[1646] [Means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal;

[1647] [Methods for analyzing customer feedback and behavioral data to propose measures to improve customer satisfaction, and

[1648] [Means for analyzing customer reviews and sentiment data in real time and making product recommendations based on related topics and sentiment scores;

[1649] [Methods for visualizing the sentiment score distribution and clustering results on a dashboard, and

[1650] A system including:

[1651] ...

Claims

1. a means for collecting information from data sources; and a means for storing the collected information in a database; a means for cleaning the stored information and formatting the data; A means to analyze the formatted data using natural language processing technology and extract important keywords and patterns; A means to train predictive models based on past data to make future predictions; A means for visualizing the prediction results and analysis results in a dashboard format and presenting them on a terminal; A means to analyze customer feedback and behavioral data and propose measures to improve customer satisfaction, A system including:

2. 10. The system of claim 1, wherein the server has means for cleaning the information collected from the data sources and analyzing it using natural language processing techniques.

3. The system according to claim 1, wherein the server trains a predictive model using past data and visualizes and presents the prediction results in a dashboard format.

Images