System

Abstract

A system is provided.SOLUTION: A system, comprising: means for collecting historical transaction data; means for analyzing the collected transaction data and extracting key phrases related to success and failure; means for generating and updating a model that predicts agreement accuracy based on the extracted key phrases; means for identifying the key phrases and making warnings or suggestions in real-time during an opportunity; and means for collecting results of the opportunity as feedback to improve accuracy of the model.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 sales activities, there is a problem in that repeated useless approaches are made in response to customers' standard phrases such as "I'll consider it," resulting in a significant drop in sales efficiency. Furthermore, it is difficult to see the factors that lead to successful and unsuccessful sales negotiations, and sales representatives often repeat the same mistakes. There is a need for a method to reduce such useless approaches and achieve rational and efficient sales activities. [Means for solving the problem]

[0005] The present invention provides a means for collecting past transaction data and a means for analyzing the collected transaction data to extract key phrases related to success and failure. It also includes a means for generating and updating a model that predicts the likelihood of agreement based on the extracted key phrases, and a means for identifying key phrases in real time during sales negotiations and issuing warnings or suggestions. It also includes a means for collecting sales negotiation results as feedback and improving the accuracy of the model. This system allows sales representatives to predict the likelihood of agreement in sales negotiations in advance, making rational decisions, and reducing wasteful approaches.

[0006] "Transaction Data" means records of information relating to past business negotiations, contracts, and agreements.

[0007] "Analysis" is the process of analyzing collected data to find specific patterns and trends.

[0008] "Success and failure" refers to whether the outcome of a business negotiation or transaction is as intended (success) or does not achieve the intended goal (failure).

[0009] A "key phrase" is a specific word or short phrase that has a significant impact on the agreement or success of a deal.

[0010] The "probability of agreement" is a predicted value that indicates the likelihood that a business deal will be concluded.

[0011] A "model" is a computational method or algorithm created to predict the likelihood of a deal being agreed based on data.

[0012] "Real time" means that the processing is currently in progress and is carried out without any time delay.

[0013] "Identification" means finding specific key phrases or patterns in the analyzed information.

[0014] "Warning" refers to a notice or signal that calls attention.

[0015] A "suggestion" is information that recommends a particular action or measure.

[0016] "Feedback" means returning information about the outcome and status of a business deal to the system.

[0017] "Accuracy" is an indicator of the accuracy and reliability of a model or prediction. [Brief explanation of the drawings]

[0018] [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

[0019] 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.

[0020] First, the terms used in the following description will be explained.

[0021] 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).

[0022] 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.

[0023] 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.

[0024] 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.

[0025] 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."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 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.

[0029] 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).

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] 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."

[0039] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0040] 1. Data Collection

[0041] The server first collects past transaction data, including meeting minutes, transaction records, and email correspondence, from the company's internal database, extracting it and converting it into text format as needed for easier analysis.

[0042] Specific examples

[0043] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0044] 2. Keyphrase Extraction

[0045] The server analyzes the collected transaction data with generative AI to extract key phrases related to the success and failure of business negotiations. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0046] Specific examples

[0047] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0048] 3. Creating and updating the model

[0049] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0050] Specific examples

[0051] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0052] 4. Real-time support for business negotiations

[0053] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0054] Specific examples

[0055] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0056] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0057] 5. Feedback and model refinement

[0058] After the sales negotiation, the user (sales representative) provides feedback to the system, and the server uses this feedback to further refine the prediction model.

[0059] Specific examples

[0060] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0061] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0062] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0063] The processing flow will be explained below.

[0064] Step 1: Data collection

[0065] The server extracts past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and converts the data into text format for analysis.

[0066] Specific actions

[0067] The server extracts the transaction data from the database every night at midnight.

[0068] The server converts the extracted data into text format and saves it.

[0069] Step 2: Extracting Keyphrases

[0070] The server analyzes the collected transaction data and extracts key phrases associated with success and failure. This analysis is performed using generative AI, employing machine learning and natural language processing (NLP) techniques.

[0071] Specific actions

[0072] The server inputs the collected transaction data into a machine learning model.

[0073] The server uses NLP technology to identify frequently occurring key phrases from a dataset of successful and unsuccessful deals.

[0074] Step 3: Generate and update the model

[0075] The server generates and updates a model for predicting the probability of a deal being concluded based on the extracted key phrases. This model includes an algorithm for predicting whether a deal will be successful or not.

[0076] Specific actions

[0077] The server uses the extracted key phrases as input data to generate regression analysis and machine learning models to predict the likelihood of agreement.

[0078] The server periodically retrains the model with new data to improve its accuracy.

[0079] Step 4: Real-time support for business negotiations

[0080] The device identifies key phrases in real time during a sales negotiation, provides warnings and suggestions as needed, and evaluates the progress of the negotiation based on the latest key phrase list provided by the server.

[0081] Specific actions

[0082] The device analyzes speech during negotiations and notifies the sales representative when key phrases appear.

[0083] The device will display a warning if important key phrases are missing from the meeting minutes.

[0084] Step 5: Feedback on the outcome of the deal

[0085] After the sales negotiation is completed, the user provides feedback on the results of the negotiation to the system, and the server uses this feedback to further refine the prediction model.

[0086] Specific actions

[0087] The user inputs information into the system about the results of the business negotiation, such as "success" or "failure."

[0088] The server analyzes the input feedback data and retrains the model to improve its predictive accuracy.

[0089] This series of steps allows salespeople to reduce unnecessary approaches and achieve rational and efficient sales activities.

[0090] Example 1

[0091] 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."

[0092] In conventional sales activities, there was insufficient data analysis to predict the success or failure of sales negotiations, and sales relied heavily on experience. As a result, there was a need for a method to efficiently and reliably increase the likelihood of sales negotiations reaching an agreement. In addition, it was difficult to provide useful information in real time during sales negotiations, which meant that sales representatives were unable to take immediate measures.

[0093] 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.

[0094] In this invention, the server includes: means for converting past transaction data into text format using optical character recognition technology; means for analyzing transaction data using a generative AI model and inputting prompt sentences; means for analyzing collected transaction data and extracting key phrases related to success and failure; means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases; means for identifying the key phrases in real time during a sales negotiation and issuing warnings or suggestions; and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to identify important key phrases that lead to the success of a sales negotiation in real time and provide immediate countermeasures. Furthermore, the model can be updated based on feedback data to continuously improve the likelihood of agreement in sales negotiations.

[0095] "Transactional data" refers to data related to past business activities, such as minutes of business meetings, transaction records, and email correspondence.

[0096] "Optical character recognition technology" refers to the technology that converts non-text documents such as images and PDFs into text data by machine.

[0097] A "generative AI model" refers to an artificial intelligence model that uses machine learning techniques to make predictions and analyze specific input data.

[0098] A "prompt sentence" refers to a sentence in the form of input that allows a generative AI model to respond appropriately.

[0099] "Key phrases" refer to important phrases or words extracted from text data that are related to the success or failure of a business deal.

[0100] "Agreement probability" refers to the degree to which the probability of a business negotiation reaching an agreement is predicted.

[0101] A "model" refers to a computational method or algorithm constructed to make predictions or perform analysis on specific input data.

[0102] "Real-time" refers to data being processed and analyzed almost as soon as it is generated.

[0103] "Feedback" refers to entering evaluations and information about the results and process of sales negotiations into the system and reflecting them in the next forecast or analysis.

[0104] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0105] Data collection

[0106] The server first accesses the company's internal database to collect past transaction data, including meeting minutes, transaction records, and email correspondence. Because this data is often stored in PDF or image format, the server converts it into text using optical character recognition technology. For example, Tesseract OCR software is used to convert PDF files into text files, which are then stored in the database.

[0107] Keyphrase Extraction

[0108] The server inputs the collected text data into a generative AI model. The generative AI model (e.g., GPT-3) uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. In this process, only key phrases with a confidence level above a certain probability are saved as a list. An example of a prompt sentence is "Please extract important key phrases that distinguish between success and failure in business negotiation activities."

[0109] Generating and Updating Models

[0110] The server uses the extracted key phrases to generate a model that predicts the likelihood of agreement. It uses machine learning algorithms (e.g., random forests or regression analysis) to learn patterns with a high probability of success from past sales negotiation data. The model is periodically updated with new data. For example, it performs regression analysis based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0111] Real-time support for business negotiations

[0112] The terminal (salesperson's device) monitors the sales negotiation record in real time during the negotiation and identifies key phrases. Based on the key phrase list provided by the server, it gives warnings and suggestions to the salesperson depending on the progress of the negotiation. For example, when the key phrase "customer support" is detected, the terminal displays a pop-up notification and makes a suggestion such as "This key phrase will increase the probability of agreement."

[0113] Feedback and model refinement

[0114] After a sales meeting is completed, the user (sales representative) provides feedback to the system on the results. They enter whether the meeting was a success or failure, and record the details of each case. The server uses this feedback to retrain the prediction model and improve prediction accuracy. For example, if a user enters feedback such as "This meeting was successful" after the meeting is completed, the server analyzes that data and the list of key phrases detected during the meeting to retrain the prediction model.

[0115] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0116] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0117] Step 1:

[0118] The server accesses the company's internal database and collects past transaction data. The transaction data includes meeting minutes, transaction records, and email correspondence. The server extracts this data from the database using SQL queries. For example, the server executes the following query: SELECT FROM meeting data WHERE date BETWEEN '2022-01-01' AND '2023-01-01'. The extracted data is converted to text format. Meeting minutes in PDF format are converted to text using OCR technology. The input is a PDF file, and the output is a text file.

[0119] Specific behavior:

[0120] The server connects to the database and extracts the minutes of business meetings from the past year using an SQL query.

[0121] After extraction, the PDF files are converted to text files using Tesseract OCR software.

[0122] The converted text file is formatted and saved in a unified format.

[0123] Step 2:

[0124] The server inputs the collected text data into a generative AI model (e.g., GPT-3). The generative AI model uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. The input prompt is "Please extract important key phrases that distinguish between success and failure in business negotiation activities," and the model outputs key phrases based on this. The input is text data, and the output is a list of key phrases.

[0125] Specific behavior:

[0126] The server inputs text data into the generative AI model.

[0127] Send the prompt "Please extract the key phrases that distinguish between success and failure in sales negotiation activities" to the model.

[0128] The generative AI model identifies key phrases and outputs them in a list format.

[0129] The server stores the extracted key phrases in a database.

[0130] Step 3:

[0131] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This process uses a machine learning algorithm (e.g., random forest or regression analysis). The input is the list of key phrases and past sales performance data, and the output is the predictive model.

[0132] Specific behavior:

[0133] The server performs regression analysis using sales negotiation data with a success rate of 80% or higher.

[0134] It learns the relationship between the frequency of key phrases and the success rate of sales negotiations, and generates a model that predicts the probability of agreement.

[0135] The trained model is stored on the server and periodically updated with new data.

[0136] Step 4:

[0137] The terminal monitors data entered in real time during sales negotiations and identifies key phrases. Based on the key phrase list provided by the server, it issues warnings or suggestions to sales representatives according to the progress of the negotiations. The input is real-time sales negotiation records, and the output is warning or suggestion notifications.

[0138] Specific behavior:

[0139] The device monitors conversations and text input during business negotiations in real time.

[0140] If the key phrase "customer support" is identified, the device will display a pop-up notification.

[0141] If an important keyphrase is missing, the terminal will display a warning message.

[0142] Step 5:

[0143] After a sales negotiation is completed, the user provides feedback to the system on the results, entering whether the negotiation was successful or not and adding detailed comments. The server uses this feedback data to retrain the prediction model and improve its accuracy. The input is the feedback data, and the output is an updated prediction model.

[0144] Specific behavior:

[0145] After the negotiation is completed, the user inputs feedback on the success of the negotiation into the system.

[0146] The server analyzes the feedback data and the key phrase data detected during the business negotiation.

[0147] Retrain and update the predictive model.

[0148] The updated model will be applied to the next deal.

[0149] In this way, this system provides support to improve the likelihood of agreement in business negotiations through each step.

[0150] (Application example 1)

[0151] 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."

[0152] Conventional sales negotiation support systems for sales activities are limited to methods that utilize past transaction data to predict the likelihood of success of negotiations, making them difficult to apply to other business environments. Furthermore, there are limitations to efficient product picking methods at logistics facilities, creating a need for improved work efficiency. Currently, there is a lack of means to make specific suggestions for optimizing work efficiency in real time during picking work. This means that optimal support for workers and robots to work efficiently cannot be provided.

[0153] 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.

[0154] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting agreement probability based on the extracted key phrases, means for identifying the key phrases in real time during negotiations and issuing warnings or suggestions, means for collecting the results of the negotiations as feedback and improving the accuracy of the model, and means for applying the model to support operational efficiency in logistics facilities and making suggestions in real time when a robot performs product picking work. This makes it possible to optimize the efficiency of picking work in logistics facilities in addition to sales activities in real time.

[0155] "Transaction data" refers to data related to sales activities, including minutes of business negotiations, transaction records, email exchanges, etc.

[0156] "Key phrases" are specific important words or phrases related to the success or failure of a business deal.

[0157] "Agreement probability" is an indicator that indicates the probability that a business negotiation will be successful.

[0158] A "model" is a computational method or algorithm used to analyze collected data and predict the likelihood of agreement.

[0159] A "logistics facility" is a facility for storing goods and for picking and shipping operations.

[0160] "Product picking work" refers to the work of picking out products at a logistics facility, packaging them, and shipping them.

[0161] A "robot" is a mechanical device that automates and efficiently performs product picking tasks at logistics facilities.

[0162] "Real-time" is a concept that refers to immediate response during negotiations or product picking.

[0163] A "proposal" is an act of giving advice or instructions to efficiently advance business negotiations or picking work.

[0164] "Feedback" refers to reflecting the results of business negotiations and picking operations in the system.

[0165] "Model accuracy" is an indicator of how closely the results predicted by the model match the actual results.

[0166] The present invention relates to a system for improving the efficiency of product picking operations in sales activities and logistics facilities. Specific embodiments of the system will be described below.

[0167] 1. Data Collection

[0168] The server first collects past transaction data. This data includes minutes of business meetings, transaction records, email correspondence, and more. It also collects past picking data from logistics facilities. The server extracts this data from the company's internal database and converts it into text format as needed, making it easier to analyze.

[0169] 2. Keyphrase Extraction

[0170] The server uses generative AI to analyze the collected transaction data and extract key phrases related to the success or failure of sales negotiations. Similarly, it extracts key phrases related to efficient work from the picking data. This analysis uses machine learning and natural language processing (NLP) techniques. Specifically, it uses TfidfVectorizer and RandomForestClassifier.

[0171] 3. Creating and updating the model

[0172] The server generates and updates a model that predicts the likelihood of a negotiation agreement and the efficiency of picking work based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of negotiations and picking work.

[0173] 4. Real-time support for sales negotiations and picking

[0174] The terminal identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as necessary. It evaluates the progress of negotiations and picking work based on a key phrase list provided by the server. For example, if the key phrase "ROI (return on investment)" appears during negotiations, the sales representative is notified that the likelihood of agreement is increasing. Also, if the key phrase "high-demand product" appears during picking work, an efficient picking order is suggested based on that.

[0175] 5. Feedback and model refinement

[0176] After a sales negotiation or picking task is completed, the user provides feedback to the system. The server uses this feedback to further refine the prediction model. For example, the user can input feedback such as "success" or "failure" as the outcome of the negotiation. The server analyzes this feedback data and retrains the model to improve its prediction accuracy.

[0177] As a specific example, the system identifies key phrases such as "high-demand products" and "high-priority items" and suggests an efficient picking order based on them.

[0178] Example prompts to input to a generative AI model:

[0179] "Please analyze past picking work logs, extract key phrases to improve picking efficiency, and suggest the optimal picking order in real time."

[0180] This system makes it possible to optimize the efficiency of sales activities as well as picking operations at logistics facilities in real time.

[0181] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0182] Step 1:

[0183] The server collects past transaction data and picking data. It extracts data from the transaction database and picking database as input and obtains data converted into text format as output. Specifically, it converts data such as minutes, transaction records, email correspondence, and picking logs into a format that is easy to centrally manage.

[0184] Step 2:

[0185] The server analyzes the collected transaction and picking data to extract key phrases related to success and failure. It uses the collected text data as input and generates a list of key phrases as output. Specifically, it vectorizes the text data using TfidfVectorizer and identifies important key phrases using RandomForestClassifier.

[0186] Step 3:

[0187] The server generates and updates models that predict the likelihood of agreement in negotiations and the efficiency of picking work based on the extracted key phrases. It uses the key phrase list as input and generates a predictive model as output. Specifically, it analyzes the frequency and relationships of key phrases and builds the model using regression analysis and machine learning algorithms.

[0188] Step 4:

[0189] The device identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as needed. It uses real-time negotiation voice data or picking logs as input and generates warning or suggestion messages as output. Specifically, it works in conjunction with voice recognition software (e.g., Google Speech-to-Text) to detect pre-defined key phrases in real time.

[0190] Step 5:

[0191] After a sales negotiation or picking task is completed, the user provides feedback to the system. The results of the negotiation (success or failure) or the results of the picking task (achievement rate, number of errors) are input to the system, and feedback data is generated as output. Specifically, the results are entered through a mobile application or web interface and sent to the server as feedback data.

[0192] Step 6:

[0193] The server retrains the predictive model based on the feedback to improve its accuracy. It uses the feedback data as input and generates an improved predictive model as output. Specifically, it feeds the newly collected feedback data into a machine learning algorithm to optimize the model's parameters.

[0194] 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.

[0195] The present invention relates to a system that uses a generation AI and an emotion engine to systematize know-how for reaching agreements in sales negotiations, and specific embodiments thereof are described below.

[0196] 1. Data Collection

[0197] The server first collects past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and then converts the data into text format for analysis.

[0198] Specific examples

[0199] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0200] 2. Keyphrase Extraction

[0201] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0202] Specific examples

[0203] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0204] 3. Creating and updating the model

[0205] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0206] Specific examples

[0207] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0208] 4. Real-time support for business negotiations

[0209] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0210] Specific examples

[0211] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0212] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0213] 5. Leveraging Emotional Engines

[0214] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0215] Specific examples

[0216] The device uses a camera to analyze the customer's facial expressions during sales negotiations and notifies the sales representative if emotions such as anxiety or excitement are detected.

[0217] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction and makes suggestions to adjust the progress of the business negotiations.

[0218] 6. Feedback on sales results

[0219] After the sales negotiation, the user (salesperson) provides feedback to the system, and the server uses this feedback to further refine the prediction model and emotion recognition model.

[0220] Specific examples

[0221] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0222] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0223] This system allows salespeople to grasp important key phrases and the emotional state of the customer during the first sales meeting, increasing the likelihood of an agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0224] The processing flow will be explained below.

[0225] Step 1: Data collection

[0226] The server extracts past transaction data from the company's internal database, including business meeting minutes, transaction records, and email correspondence. The collected data is then converted into text format, making it suitable for analysis.

[0227] Specific actions

[0228] The server automatically extracts transaction data from the database every night.

[0229] The server converts the extracted data into text format and prepares it into an analyzable file format.

[0230] Step 2: Extracting Keyphrases

[0231] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis utilizes machine learning and natural language processing (NLP) techniques.

[0232] Specific actions

[0233] The server inputs the collected transaction data into a machine learning model and performs NLP analysis.

[0234] The server compares data from successful and unsuccessful transactions to identify frequently occurring key phrases.

[0235] Step 3: Generate and update the model

[0236] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically.

[0237] Specific actions

[0238] The server creates a predictive model using regression analysis and machine learning algorithms based on key phrases that frequently appear in business negotiations with a high success rate.

[0239] The server periodically retrains the model with newly collected data to improve prediction accuracy.

[0240] Step 4: Real-time support for business negotiations

[0241] The terminal, as a device for sales representatives, identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the latest key phrase list provided by the server.

[0242] Specific actions

[0243] The device uses voice recognition technology to analyze speech during negotiations and notify the sales representative when important key phrases appear.

[0244] The terminal displays a warning if an important key phrase is missing when entering business meeting minutes.

[0245] Step 5: Leverage the Emotion Engine

[0246] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0247] Specific actions

[0248] The device uses a camera to analyze the customer's facial expressions during the sales negotiation and recognizes their emotional state (e.g., anxiety, excitement, satisfaction, etc.) in real time.

[0249] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction, providing the sales representative with information to adjust the progress of the negotiation and the content of the proposal.

[0250] Step 6: Feedback on the outcome of the deal

[0251] After a sales meeting, the user (sales representative) provides feedback to the system, and the server uses this feedback to improve the accuracy of the prediction model and emotion engine.

[0252] Specific actions

[0253] The user enters the outcome of the negotiation (success or failure) and the reason for it into the system.

[0254] The server analyzes the feedback data and retrains the predictive model and emotion engine to improve accuracy.

[0255] Example 2

[0256] 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."

[0257] To increase the likelihood of agreement in sales negotiations, it is important to utilize past transaction data to identify factors related to success and failure. However, conventional methods make it difficult to efficiently analyze these factors and support sales negotiations in real time. In addition, more effective sales negotiation support is required by understanding the customer's emotional state, but there is a lack of technology to perform emotional analysis in real time. To solve these issues, a comprehensive and efficient system is required.

[0258] 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.

[0259] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases, means for identifying key phrases in real time during a sales negotiation and issuing a warning or suggestion, means for analyzing the emotional state of the customer in real time using a sentiment analysis engine and notifying the sales representative, and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to accurately grasp important factors in sales negotiations and increase the likelihood of agreement in sales negotiations.

[0260] "Transaction data" refers to a series of records related to business negotiations, such as minutes of past business negotiations, transaction records, and email exchanges.

[0261] "Key phrases" refer to important words or short phrases related to the success or failure of a business deal, extracted through analysis of transaction data.

[0262] An "agreement probability prediction model" refers to a mathematical model that includes calculation methods and algorithms for predicting the success or failure of a business negotiation based on extracted key phrases.

[0263] "Generative AI" refers to artificial intelligence that uses machine learning and natural language processing techniques to analyze collected transaction data and extract key phrases.

[0264] "Feedback" refers to the information on the results of sales negotiations that sales representatives enter into the system after the negotiations are completed, and is data used to improve the accuracy of the model.

[0265] An "emotion analysis engine" refers to technology that analyzes data such as a customer's tone of voice and facial expressions to recognize their emotional state in real time.

[0266] "Real-time analysis" refers to a processing method that processes data instantly while a sales negotiation is in progress and provides sales representatives with the necessary information and warnings immediately.

[0267] "Salesperson's device" refers to a device (e.g., PC, tablet, smartphone) that receives real-time information and displays alerts and suggestions during a sales meeting.

[0268] The present invention relates to a system for systematizing sales negotiation know-how using a generation AI and a sentiment analysis engine. Specific embodiments of the system are described below.

[0269] First, the server collects past transaction data from the company's internal database. This transaction data includes sales meeting minutes, transaction records, and email correspondence, and converts it into text format and saves it. For example, the server uses an SQL query to extract sales data from January 2022 to January 2023, and then uses a Python script to save it as a text JSON file.

[0270] Next, the server analyzes the collected transaction data using a generative AI (e.g., GPT-4) to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques. Specifically, the server preprocesses the collected text data using a natural language processing library (e.g., spaCy, NLTK), inputs it into the generative AI to extract key phrases, and stores the extracted key phrases in a database.

[0271] The server then generates and updates a model to predict the likelihood of agreement based on the extracted key phrases. This model includes algorithms (e.g., regression analysis, machine learning) for predicting the success or failure of a deal. Specifically, the server combines the extracted key phrases with the deal outcome data, trains a regression model or classifier using a machine learning library (e.g., scikit-learn), and stores the trained model in a model repository on the server.

[0272] The terminal (sales representative's device) identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The progress of the negotiation is evaluated based on the key phrase list provided by the server. Specifically, the terminal uses a real-time data analysis module to convert the conversation during the negotiation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears. It also issues a warning if an important key phrase is omitted when creating minutes of the negotiation.

[0273] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The emotion analysis engine recognizes emotions by analyzing data points such as the customer's voice tone and facial expressions. Specifically, the device captures the customer's facial expressions with a camera, performs image analysis using a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0274] Finally, after the sales negotiation, the user (salesperson) provides feedback to the system on the outcome. The server uses this feedback to further refine the prediction model and emotion recognition model. Specifically, the user enters "success" or "failure" as the outcome of the negotiation into the system, and the server uses this feedback data as a new learning dataset to retrain the model.

[0275] Example prompt

[0276] 1. "Extract key phrases related to success and failure from past transaction data."

[0277] 2. "Suggest a way to analyze a customer's tone of voice and facial expressions to signal their emotional state during a sales conversation."

[0278] 3. "How can we refine our predictive models using data from sales rep feedback on deal outcomes?"

[0279] By utilizing this system, sales representatives can accurately grasp the important elements of sales negotiations, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0280] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0281] Step 1:

[0282] The server collects past transaction data from the company's internal database. Input data includes meeting minutes, transaction records, and email correspondence. This data is converted into text format and output. Specifically, the server uses SQL queries to extract the transaction data, converts it into a text-format JSON file using a Python script, and saves it.

[0283] Step 2:

[0284] The server analyzes the collected transaction data using a generation AI and extracts key phrases related to success and failure. The input is text-format transaction data, and the output is the extracted key phrases. Specifically, the server performs preprocessing using a natural language processing library (e.g., spaCy, NLTK), inputs the data into the generation AI to extract key phrases, and stores them in a database.

[0285] Step 3:

[0286] The server generates and updates a model that predicts the probability of agreement based on the extracted key phrases. The inputs are key phrases and negotiation outcome data, and the output is a predictive model. Specifically, the server creates a training dataset using a machine learning library (e.g., scikit-learn), trains a regression model or classifier, and saves the trained model in a model repository.

[0287] Step 4:

[0288] The terminal identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The input is conversation data from the negotiation, and the output is the identified key phrases and warnings or suggestions based on them. Specifically, the terminal uses a real-time data analysis module to convert the conversation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears.

[0289] Step 5:

[0290] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The inputs are voice tone and facial expression data, and the output is the analyzed emotional state. Specifically, the device captures the customer's facial expression with a camera, analyzes it with a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0291] Step 6:

[0292] After a sales negotiation, the user (salesperson) feeds the results back to the system. The input is the negotiation result (success or failure), and the output is an updated prediction model. Specifically, the user inputs the negotiation result into the system, and the server uses this feedback data to retrain the model, and retrains it to improve the model's accuracy.

[0293] (Application example 2)

[0294] 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."

[0295] The progress of sales negotiations and consensus building in sales activities often depend on the experience and intuition of individual sales representatives, resulting in variations in the success rate of sales negotiations. Furthermore, it is not easy to appropriately grasp and respond to the customer's emotional state, and effective methods are needed to increase the likelihood of agreement in sales negotiations. Furthermore, systems that support sales negotiations in real time lack accuracy and practicality, so the development of more advanced and effective sales negotiation support systems is necessary.

[0296] 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.

[0297] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data to extract key phrases related to success and failure, and means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases. This allows sales representatives to identify the key phrases in real time during negotiations and receive warnings or suggestions. Furthermore, by including emotion analysis means for analyzing the emotional state of the customer and means for making suggestions for adjusting the progress of the negotiations based on the emotional data analyzed by the emotion analysis means, appropriate responses can be made based on the customer's emotions, further increasing the likelihood of agreement in the negotiations. Furthermore, by including means for collecting the results of the negotiations as feedback and improving the accuracy of the model, the accuracy of the system can be continuously improved.

[0298] "Transaction data" refers to all information related to past business negotiations and negotiations, including minutes of business negotiations, transaction records, and email correspondence.

[0299] "Collection means" is a general term for the equipment and programs used to compile the necessary data and gather information.

[0300] "Analysis tools" is a general term for the equipment and algorithms used to examine collected data in detail and extract meaningful information from it.

[0301] "Key phrases" refer to important phrases or terms related to the success or failure of a deal.

[0302] "Agreement likelihood" refers to the likelihood that a particular deal will be successful and result in an agreement.

[0303] "Emotion analysis means" is a general term for devices and programs that analyze a customer's tone of voice, facial expressions, movements, etc., to identify the customer's emotional state.

[0304] "Real-time identification means" is a general term for devices and programs that perform real-time data analysis during business negotiations and instantly identify important information.

[0305] "Feedback tools" is a general term for devices and programs that systematically collect results obtained after business negotiations and are used to improve and learn from the system.

[0306] This invention is a system that uses a generation AI and an emotion engine to systematize sales negotiation agreement know-how in sales activities, and specific embodiments will be described below.

[0307] 1. Data Collection and Preprocessing

[0308] The server first collects data on past business negotiations and transactions from the company's internal database, including minutes of business negotiations, transaction records, and email correspondence, and then converts the collected data into text format for analysis.

[0309] 2. Keyphrase Extraction

[0310] The server analyzes the collected transaction data using generative AI to extract key phrases related to success and failure. This analysis uses machine learning and natural language processing (NLP) techniques. Specific tools include Hugging Face's Transformers.

[0311] 3. Creating and updating the model

[0312] The server generates and periodically updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model uses regression analysis or machine learning algorithms, such as open-source libraries like scikit-learn.

[0313] 4. Real-time support for business negotiations

[0314] The salesperson's device (e.g., smart glasses) identifies key phrases in real time during a sales call and provides warnings or suggestions as needed. The device immediately notifies the customer when key phrases such as "warranty period" or "additional services" appear during the call. Software such as OpenCV and SpeechRecognition are also used to analyze the customer's facial expressions and tone of voice and suggest appropriate responses based on their emotional state.

[0315] 5. Leveraging Emotional Engines

[0316] The device uses an emotion engine to analyze the customer's emotional state and provide feedback to the salesperson. For example, if the customer looks anxious, the device will use voice tone analysis to suggest appropriate countermeasures.

[0317] 6. Feedback on sales results

[0318] After a sales meeting is completed, the sales representative provides feedback to the system, and the server uses this feedback to further improve the accuracy of the prediction model and emotion recognition model.

[0319] Specific examples

[0320] For example, if a salesperson wearing smart glasses is asked "How long is the warranty period?" during a sales negotiation with a customer in a physical store, the system will detect the important key phrase "warranty period." If the customer looks anxious, the system will notify the salesperson to provide a more detailed explanation, thereby helping to increase the likelihood of an agreement.

[0321] Prompt Sentence Examples

[0322] When a customer asks, "How long is the product warranty period?", the system detects the important key phrase "warranty period" and notifies the sales representative. If anxiety is detected from the customer's facial expression, the system suggests a solution.

[0323] This allows salespeople to respond optimally to the needs and emotional state of the customer, thereby improving the success rate of sales negotiations.

[0324] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0325] Flow of the system program that realizes the application example

[0326] Step 1:

[0327] The server collects past transaction data from the company's internal database and converts it into text format. The collected data includes minutes of business meetings, transaction records, and email correspondence. It receives data from the company's internal database as input and outputs text data that can be analyzed.

[0328] Step 2:

[0329] The server analyzes the collected text data using generative AI and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. It receives text data as input and obtains key phrases related to success and failure as output. This processing uses Hugging Face's Transformers library.

[0330] Step 3:

[0331] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically. Specifically, it uses regression analysis and machine learning algorithms. It receives key phrases as input and obtains a predictive model as output. Machine learning libraries such as scikit-learn are used.

[0332] Step 4:

[0333] The device (e.g., smart glasses) captures audio and video in real time during a sales meeting and analyzes it. Specifically, the camera captures the customer's facial expressions and the microphone records their voice. The real-time data obtained from the camera and microphone is received as input, and the analysis results are obtained as output. OpenCV and the SpeechRecognition library are used.

[0334] Step 5:

[0335] The device uses the generative AI model from the server to identify key phrases that appear during sales negotiations in real time and provide warnings or suggestions to the sales representative as needed. It receives real-time analysis results and predictive models as input and provides notifications and suggestions as output.

[0336] Step 6:

[0337] The device uses an emotion analysis engine to analyze the voice tone and facial expression data during the negotiation to detect the customer's emotional state. It receives audio and video data as input and outputs the analysis result of the customer's emotional state.

[0338] Step 7:

[0339] The device then proposes appropriate countermeasures to the salesperson based on the results of the emotion analysis. For example, if the customer is feeling anxious, it will notify them to provide a detailed explanation. The device receives the emotion analysis results as input and proposes countermeasures as output.

[0340] Step 8:

[0341] After the negotiation is completed, the user provides feedback on the outcome (success or failure) to the system, and the server retrains the prediction model based on this data. The server receives the negotiation result data as input and obtains an improved prediction model as output.

[0342] 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.

[0343] 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.

[0344] 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.

[0345] [Second embodiment]

[0346] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0347] 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.

[0348] 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).

[0349] 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.

[0350] 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.

[0351] 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).

[0352] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[0353] 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.

[0354] 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.

[0355] 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.

[0356] 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.

[0357] 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."

[0358] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0359] 1. Data Collection

[0360] The server first collects past transaction data, including meeting minutes, transaction records, and email correspondence, from the company's internal database, extracting it and converting it into text format as needed for easier analysis.

[0361] Specific examples

[0362] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0363] 2. Keyphrase Extraction

[0364] The server analyzes the collected transaction data with generative AI to extract key phrases related to the success and failure of business negotiations. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0365] Specific examples

[0366] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0367] 3. Creating and updating the model

[0368] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0369] Specific examples

[0370] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0371] 4. Real-time support for business negotiations

[0372] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0373] Specific examples

[0374] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0375] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0376] 5. Feedback and model refinement

[0377] After the sales negotiation, the user (sales representative) provides feedback to the system, and the server uses this feedback to further refine the prediction model.

[0378] Specific examples

[0379] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0380] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0381] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0382] The processing flow will be explained below.

[0383] Step 1: Data collection

[0384] The server extracts past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and converts the data into text format for analysis.

[0385] Specific actions

[0386] The server extracts the transaction data from the database every night at midnight.

[0387] The server converts the extracted data into text format and saves it.

[0388] Step 2: Extracting Keyphrases

[0389] The server analyzes the collected transaction data and extracts key phrases associated with success and failure. This analysis is performed using generative AI, employing machine learning and natural language processing (NLP) techniques.

[0390] Specific actions

[0391] The server inputs the collected transaction data into a machine learning model.

[0392] The server uses NLP technology to identify frequently occurring key phrases from a dataset of successful and unsuccessful deals.

[0393] Step 3: Generate and update the model

[0394] The server generates and updates a model for predicting the probability of a deal being concluded based on the extracted key phrases. This model includes an algorithm for predicting whether a deal will be successful or not.

[0395] Specific actions

[0396] The server uses the extracted key phrases as input data to generate regression analysis and machine learning models to predict the likelihood of agreement.

[0397] The server periodically retrains the model with new data to improve its accuracy.

[0398] Step 4: Real-time support for business negotiations

[0399] The device identifies key phrases in real time during a sales negotiation, provides warnings and suggestions as needed, and evaluates the progress of the negotiation based on the latest key phrase list provided by the server.

[0400] Specific actions

[0401] The device analyzes speech during negotiations and notifies the sales representative when key phrases appear.

[0402] The device will display a warning if important key phrases are missing from the meeting minutes.

[0403] Step 5: Feedback on the outcome of the deal

[0404] After the sales negotiation is completed, the user provides feedback on the results of the negotiation to the system, and the server uses this feedback to further refine the prediction model.

[0405] Specific actions

[0406] The user inputs information into the system about the results of the business negotiation, such as "success" or "failure."

[0407] The server analyzes the input feedback data and retrains the model to improve its predictive accuracy.

[0408] This series of steps allows salespeople to reduce unnecessary approaches and achieve rational and efficient sales activities.

[0409] Example 1

[0410] 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."

[0411] In conventional sales activities, there was insufficient data analysis to predict the success or failure of sales negotiations, and sales relied heavily on experience. As a result, there was a need for a method to efficiently and reliably increase the likelihood of sales negotiations reaching an agreement. In addition, it was difficult to provide useful information in real time during sales negotiations, which meant that sales representatives were unable to take immediate measures.

[0412] 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.

[0413] In this invention, the server includes: means for converting past transaction data into text format using optical character recognition technology; means for analyzing transaction data using a generative AI model and inputting prompt sentences; means for analyzing collected transaction data and extracting key phrases related to success and failure; means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases; means for identifying the key phrases in real time during a sales negotiation and issuing warnings or suggestions; and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to identify important key phrases that lead to the success of a sales negotiation in real time and provide immediate countermeasures. Furthermore, the model can be updated based on feedback data to continuously improve the likelihood of agreement in sales negotiations.

[0414] "Transactional data" refers to data related to past business activities, such as minutes of business meetings, transaction records, and email correspondence.

[0415] "Optical character recognition technology" refers to the technology that converts non-text documents such as images and PDFs into text data by machine.

[0416] A "generative AI model" refers to an artificial intelligence model that uses machine learning techniques to make predictions and analyze specific input data.

[0417] A "prompt sentence" refers to a sentence in the form of input that allows a generative AI model to respond appropriately.

[0418] "Key phrases" refer to important phrases or words extracted from text data that are related to the success or failure of a business deal.

[0419] "Agreement probability" refers to the degree to which the probability of a business negotiation reaching an agreement is predicted.

[0420] A "model" refers to a computational method or algorithm constructed to make predictions or perform analysis on specific input data.

[0421] "Real-time" refers to data being processed and analyzed almost as soon as it is generated.

[0422] "Feedback" refers to entering evaluations and information about the results and process of sales negotiations into the system and reflecting them in the next forecast or analysis.

[0423] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0424] Data collection

[0425] The server first accesses the company's internal database to collect past transaction data, including meeting minutes, transaction records, and email correspondence. Because this data is often stored in PDF or image format, the server converts it into text using optical character recognition technology. For example, Tesseract OCR software is used to convert PDF files into text files, which are then stored in the database.

[0426] Keyphrase Extraction

[0427] The server inputs the collected text data into a generative AI model. The generative AI model (e.g., GPT-3) uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. In this process, only key phrases with a confidence level above a certain probability are saved as a list. An example of a prompt sentence is "Please extract important key phrases that distinguish between success and failure in business negotiation activities."

[0428] Generating and Updating Models

[0429] The server uses the extracted key phrases to generate a model that predicts the likelihood of agreement. It uses machine learning algorithms (e.g., random forests or regression analysis) to learn patterns with a high probability of success from past sales negotiation data. The model is periodically updated with new data. For example, it performs regression analysis based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0430] Real-time support for business negotiations

[0431] The terminal (salesperson's device) monitors the sales negotiation record in real time during the negotiation and identifies key phrases. Based on the key phrase list provided by the server, it gives warnings and suggestions to the salesperson depending on the progress of the negotiation. For example, when the key phrase "customer support" is detected, the terminal displays a pop-up notification and makes a suggestion such as "This key phrase will increase the probability of agreement."

[0432] Feedback and model refinement

[0433] After a sales meeting is completed, the user (sales representative) provides feedback to the system on the results. They enter whether the meeting was a success or failure, and record the details of each case. The server uses this feedback to retrain the prediction model and improve prediction accuracy. For example, if a user enters feedback such as "This meeting was successful" after the meeting is completed, the server analyzes that data and the list of key phrases detected during the meeting to retrain the prediction model.

[0434] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0435] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0436] Step 1:

[0437] The server accesses the company's internal database and collects past transaction data. The transaction data includes meeting minutes, transaction records, and email correspondence. The server extracts this data from the database using SQL queries. For example, the server executes the following query: SELECT FROM meeting data WHERE date BETWEEN '2022-01-01' AND '2023-01-01'. The extracted data is converted to text format. Meeting minutes in PDF format are converted to text using OCR technology. The input is a PDF file, and the output is a text file.

[0438] Specific behavior:

[0439] The server connects to the database and extracts the minutes of business meetings from the past year using an SQL query.

[0440] After extraction, the PDF files are converted to text files using Tesseract OCR software.

[0441] The converted text file is formatted and saved in a unified format.

[0442] Step 2:

[0443] The server inputs the collected text data into a generative AI model (e.g., GPT-3). The generative AI model uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. The input prompt is "Please extract important key phrases that distinguish between success and failure in business negotiation activities," and the model outputs key phrases based on this. The input is text data, and the output is a list of key phrases.

[0444] Specific behavior:

[0445] The server inputs text data into the generative AI model.

[0446] Send the prompt "Please extract the key phrases that distinguish between success and failure in sales negotiation activities" to the model.

[0447] The generative AI model identifies key phrases and outputs them in a list format.

[0448] The server stores the extracted key phrases in a database.

[0449] Step 3:

[0450] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This process uses a machine learning algorithm (e.g., random forest or regression analysis). The input is the list of key phrases and past sales performance data, and the output is the predictive model.

[0451] Specific behavior:

[0452] The server performs regression analysis using sales negotiation data with a success rate of 80% or higher.

[0453] It learns the relationship between the frequency of key phrases and the success rate of sales negotiations, and generates a model that predicts the probability of agreement.

[0454] The trained model is stored on the server and periodically updated with new data.

[0455] Step 4:

[0456] The terminal monitors data entered in real time during sales negotiations and identifies key phrases. Based on the key phrase list provided by the server, it issues warnings or suggestions to sales representatives according to the progress of the negotiations. The input is real-time sales negotiation records, and the output is warning or suggestion notifications.

[0457] Specific behavior:

[0458] The device monitors conversations and text input during business negotiations in real time.

[0459] If the key phrase "customer support" is identified, the device will display a pop-up notification.

[0460] If an important keyphrase is missing, the terminal will display a warning message.

[0461] Step 5:

[0462] After a sales negotiation is completed, the user provides feedback to the system on the results, entering whether the negotiation was successful or not and adding detailed comments. The server uses this feedback data to retrain the prediction model and improve its accuracy. The input is the feedback data, and the output is an updated prediction model.

[0463] Specific behavior:

[0464] After the negotiation is completed, the user inputs feedback on the success of the negotiation into the system.

[0465] The server analyzes the feedback data and the key phrase data detected during the business negotiation.

[0466] Retrain and update the predictive model.

[0467] The updated model will be applied to the next deal.

[0468] In this way, this system provides support to improve the likelihood of agreement in business negotiations through each step.

[0469] (Application example 1)

[0470] 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."

[0471] Conventional sales negotiation support systems for sales activities are limited to methods that utilize past transaction data to predict the likelihood of success of negotiations, making them difficult to apply to other business environments. Furthermore, there are limitations to efficient product picking methods at logistics facilities, creating a need for improved work efficiency. Currently, there is a lack of means to make specific suggestions for optimizing work efficiency in real time during picking work. This means that optimal support for workers and robots to work efficiently cannot be provided.

[0472] 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.

[0473] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting agreement probability based on the extracted key phrases, means for identifying the key phrases in real time during negotiations and issuing warnings or suggestions, means for collecting the results of the negotiations as feedback and improving the accuracy of the model, and means for applying the model to support operational efficiency in logistics facilities and making suggestions in real time when a robot performs product picking work. This makes it possible to optimize the efficiency of picking work in logistics facilities in addition to sales activities in real time.

[0474] "Transaction data" refers to data related to sales activities, including minutes of business negotiations, transaction records, email exchanges, etc.

[0475] "Key phrases" are specific important words or phrases related to the success or failure of a business deal.

[0476] "Agreement probability" is an indicator that indicates the probability that a business negotiation will be successful.

[0477] A "model" is a computational method or algorithm used to analyze collected data and predict the likelihood of agreement.

[0478] A "logistics facility" is a facility for storing goods and for picking and shipping operations.

[0479] "Product picking work" refers to the work of picking out products at a logistics facility, packaging them, and shipping them.

[0480] A "robot" is a mechanical device that automates and efficiently performs product picking tasks at logistics facilities.

[0481] "Real-time" is a concept that refers to immediate response during negotiations or product picking.

[0482] A "proposal" is an act of giving advice or instructions to efficiently advance business negotiations or picking work.

[0483] "Feedback" refers to reflecting the results of business negotiations and picking operations in the system.

[0484] "Model accuracy" is an indicator of how closely the results predicted by the model match the actual results.

[0485] The present invention relates to a system for improving the efficiency of product picking operations in sales activities and logistics facilities. Specific embodiments of the system will be described below.

[0486] 1. Data Collection

[0487] The server first collects past transaction data. This data includes minutes of business meetings, transaction records, email correspondence, and more. It also collects past picking data from logistics facilities. The server extracts this data from the company's internal database and converts it into text format as needed, making it easier to analyze.

[0488] 2. Keyphrase Extraction

[0489] The server uses generative AI to analyze the collected transaction data and extract key phrases related to the success or failure of sales negotiations. Similarly, it extracts key phrases related to efficient work from the picking data. This analysis uses machine learning and natural language processing (NLP) techniques. Specifically, it uses TfidfVectorizer and RandomForestClassifier.

[0490] 3. Creating and updating the model

[0491] The server generates and updates a model that predicts the likelihood of a negotiation agreement and the efficiency of picking work based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of negotiations and picking work.

[0492] 4. Real-time support for sales negotiations and picking

[0493] The terminal identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as necessary. It evaluates the progress of negotiations and picking work based on a key phrase list provided by the server. For example, if the key phrase "ROI (return on investment)" appears during negotiations, the sales representative is notified that the likelihood of agreement is increasing. Also, if the key phrase "high-demand product" appears during picking work, an efficient picking order is suggested based on that.

[0494] 5. Feedback and model refinement

[0495] After a sales negotiation or picking task is completed, the user provides feedback to the system. The server uses this feedback to further refine the prediction model. For example, the user can input feedback such as "success" or "failure" as the outcome of the negotiation. The server analyzes this feedback data and retrains the model to improve its prediction accuracy.

[0496] As a specific example, the system identifies key phrases such as "high-demand products" and "high-priority items" and suggests an efficient picking order based on them.

[0497] Example prompts to input to a generative AI model:

[0498] "Please analyze past picking work logs, extract key phrases to improve picking efficiency, and suggest the optimal picking order in real time."

[0499] This system makes it possible to optimize the efficiency of sales activities as well as picking operations at logistics facilities in real time.

[0500] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0501] Step 1:

[0502] The server collects past transaction data and picking data. It extracts data from the transaction database and picking database as input and obtains data converted into text format as output. Specifically, it converts data such as minutes, transaction records, email correspondence, and picking logs into a format that is easy to centrally manage.

[0503] Step 2:

[0504] The server analyzes the collected transaction and picking data to extract key phrases related to success and failure. It uses the collected text data as input and generates a list of key phrases as output. Specifically, it vectorizes the text data using TfidfVectorizer and identifies important key phrases using RandomForestClassifier.

[0505] Step 3:

[0506] The server generates and updates models that predict the likelihood of agreement in negotiations and the efficiency of picking work based on the extracted key phrases. It uses the key phrase list as input and generates a predictive model as output. Specifically, it analyzes the frequency and relationships of key phrases and builds the model using regression analysis and machine learning algorithms.

[0507] Step 4:

[0508] The device identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as needed. It uses real-time negotiation voice data or picking logs as input and generates warning or suggestion messages as output. Specifically, it works in conjunction with voice recognition software (e.g., Google Speech-to-Text) to detect pre-defined key phrases in real time.

[0509] Step 5:

[0510] After a sales negotiation or picking task is completed, the user provides feedback to the system. The results of the negotiation (success or failure) or the results of the picking task (achievement rate, number of errors) are input to the system, and feedback data is generated as output. Specifically, the results are entered through a mobile application or web interface and sent to the server as feedback data.

[0511] Step 6:

[0512] The server retrains the predictive model based on the feedback to improve its accuracy. It uses the feedback data as input and generates an improved predictive model as output. Specifically, it feeds the newly collected feedback data into a machine learning algorithm to optimize the model's parameters.

[0513] 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.

[0514] The present invention relates to a system that uses a generation AI and an emotion engine to systematize know-how for reaching agreements in sales negotiations, and specific embodiments thereof are described below.

[0515] 1. Data Collection

[0516] The server first collects past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and then converts the data into text format for analysis.

[0517] Specific examples

[0518] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0519] 2. Keyphrase Extraction

[0520] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0521] Specific examples

[0522] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0523] 3. Creating and updating the model

[0524] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0525] Specific examples

[0526] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0527] 4. Real-time support for business negotiations

[0528] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0529] Specific examples

[0530] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0531] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0532] 5. Leveraging Emotional Engines

[0533] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0534] Specific examples

[0535] The device uses a camera to analyze the customer's facial expressions during sales negotiations and notifies the sales representative if emotions such as anxiety or excitement are detected.

[0536] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction and makes suggestions to adjust the progress of the business negotiations.

[0537] 6. Feedback on sales results

[0538] After the sales negotiation, the user (salesperson) provides feedback to the system, and the server uses this feedback to further refine the prediction model and emotion recognition model.

[0539] Specific examples

[0540] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0541] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0542] This system allows salespeople to grasp important key phrases and the emotional state of the customer during the first sales meeting, increasing the likelihood of an agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0543] The processing flow will be explained below.

[0544] Step 1: Data collection

[0545] The server extracts past transaction data from the company's internal database, including business meeting minutes, transaction records, and email correspondence. The collected data is then converted into text format, making it suitable for analysis.

[0546] Specific actions

[0547] The server automatically extracts transaction data from the database every night.

[0548] The server converts the extracted data into text format and prepares it into an analyzable file format.

[0549] Step 2: Extracting Keyphrases

[0550] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis utilizes machine learning and natural language processing (NLP) techniques.

[0551] Specific actions

[0552] The server inputs the collected transaction data into a machine learning model and performs NLP analysis.

[0553] The server compares data from successful and unsuccessful transactions to identify frequently occurring key phrases.

[0554] Step 3: Generate and update the model

[0555] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically.

[0556] Specific actions

[0557] The server creates a predictive model using regression analysis and machine learning algorithms based on key phrases that frequently appear in business negotiations with a high success rate.

[0558] The server periodically retrains the model with newly collected data to improve prediction accuracy.

[0559] Step 4: Real-time support for business negotiations

[0560] The terminal, as a device for sales representatives, identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the latest key phrase list provided by the server.

[0561] Specific actions

[0562] The device uses voice recognition technology to analyze speech during negotiations and notify the sales representative when important key phrases appear.

[0563] The terminal displays a warning if an important key phrase is missing when entering business meeting minutes.

[0564] Step 5: Leverage the Emotion Engine

[0565] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0566] Specific actions

[0567] The device uses a camera to analyze the customer's facial expressions during the sales negotiation and recognizes their emotional state (e.g., anxiety, excitement, satisfaction, etc.) in real time.

[0568] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction, providing the sales representative with information to adjust the progress of the negotiation and the content of the proposal.

[0569] Step 6: Feedback on the outcome of the deal

[0570] After a sales meeting, the user (sales representative) provides feedback to the system, and the server uses this feedback to improve the accuracy of the prediction model and emotion engine.

[0571] Specific actions

[0572] The user enters the outcome of the negotiation (success or failure) and the reason for it into the system.

[0573] The server analyzes the feedback data and retrains the predictive model and emotion engine to improve accuracy.

[0574] Example 2

[0575] 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."

[0576] To increase the likelihood of agreement in sales negotiations, it is important to utilize past transaction data to identify factors related to success and failure. However, conventional methods make it difficult to efficiently analyze these factors and support sales negotiations in real time. In addition, more effective sales negotiation support is required by understanding the customer's emotional state, but there is a lack of technology to perform emotional analysis in real time. To solve these issues, a comprehensive and efficient system is required.

[0577] 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.

[0578] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases, means for identifying key phrases in real time during a sales negotiation and issuing a warning or suggestion, means for analyzing the emotional state of the customer in real time using a sentiment analysis engine and notifying the sales representative, and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to accurately grasp important factors in sales negotiations and increase the likelihood of agreement in sales negotiations.

[0579] "Transaction data" refers to a series of records related to business negotiations, such as minutes of past business negotiations, transaction records, and email exchanges.

[0580] "Key phrases" refer to important words or short phrases related to the success or failure of a business deal, extracted through analysis of transaction data.

[0581] An "agreement probability prediction model" refers to a mathematical model that includes calculation methods and algorithms for predicting the success or failure of a business negotiation based on extracted key phrases.

[0582] "Generative AI" refers to artificial intelligence that uses machine learning and natural language processing techniques to analyze collected transaction data and extract key phrases.

[0583] "Feedback" refers to the information on the results of sales negotiations that sales representatives enter into the system after the negotiations are completed, and is data used to improve the accuracy of the model.

[0584] An "emotion analysis engine" refers to technology that analyzes data such as a customer's tone of voice and facial expressions to recognize their emotional state in real time.

[0585] "Real-time analysis" refers to a processing method that processes data instantly while a sales negotiation is in progress and provides sales representatives with the necessary information and warnings immediately.

[0586] "Salesperson's device" refers to a device (e.g., PC, tablet, smartphone) that receives real-time information and displays alerts and suggestions during a sales meeting.

[0587] The present invention relates to a system for systematizing sales negotiation know-how using a generation AI and a sentiment analysis engine. Specific embodiments of the system are described below.

[0588] First, the server collects past transaction data from the company's internal database. This transaction data includes sales meeting minutes, transaction records, and email correspondence, and converts it into text format and saves it. For example, the server uses an SQL query to extract sales data from January 2022 to January 2023, and then uses a Python script to save it as a text JSON file.

[0589] Next, the server analyzes the collected transaction data using a generative AI (e.g., GPT-4) to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques. Specifically, the server preprocesses the collected text data using a natural language processing library (e.g., spaCy, NLTK), inputs it into the generative AI to extract key phrases, and stores the extracted key phrases in a database.

[0590] The server then generates and updates a model to predict the likelihood of agreement based on the extracted key phrases. This model includes algorithms (e.g., regression analysis, machine learning) for predicting the success or failure of a deal. Specifically, the server combines the extracted key phrases with the deal outcome data, trains a regression model or classifier using a machine learning library (e.g., scikit-learn), and stores the trained model in a model repository on the server.

[0591] The terminal (sales representative's device) identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The progress of the negotiation is evaluated based on the key phrase list provided by the server. Specifically, the terminal uses a real-time data analysis module to convert the conversation during the negotiation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears. It also issues a warning if an important key phrase is omitted when creating minutes of the negotiation.

[0592] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The emotion analysis engine recognizes emotions by analyzing data points such as the customer's voice tone and facial expressions. Specifically, the device captures the customer's facial expressions with a camera, performs image analysis using a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0593] Finally, after the sales negotiation, the user (salesperson) provides feedback to the system on the outcome. The server uses this feedback to further refine the prediction model and emotion recognition model. Specifically, the user enters "success" or "failure" as the outcome of the negotiation into the system, and the server uses this feedback data as a new learning dataset to retrain the model.

[0594] Example prompt

[0595] 1. "Extract key phrases related to success and failure from past transaction data."

[0596] 2. "Suggest a way to analyze a customer's tone of voice and facial expressions to signal their emotional state during a sales conversation."

[0597] 3. "How can we refine our predictive models using data from sales rep feedback on deal outcomes?"

[0598] By utilizing this system, sales representatives can accurately grasp the important elements of sales negotiations, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0599] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0600] Step 1:

[0601] The server collects past transaction data from the company's internal database. Input data includes meeting minutes, transaction records, and email correspondence. This data is converted into text format and output. Specifically, the server uses SQL queries to extract the transaction data, converts it into a text-format JSON file using a Python script, and saves it.

[0602] Step 2:

[0603] The server analyzes the collected transaction data using a generation AI and extracts key phrases related to success and failure. The input is text-format transaction data, and the output is the extracted key phrases. Specifically, the server performs preprocessing using a natural language processing library (e.g., spaCy, NLTK), inputs the data into the generation AI to extract key phrases, and stores them in a database.

[0604] Step 3:

[0605] The server generates and updates a model that predicts the probability of agreement based on the extracted key phrases. The inputs are key phrases and negotiation outcome data, and the output is a predictive model. Specifically, the server creates a training dataset using a machine learning library (e.g., scikit-learn), trains a regression model or classifier, and saves the trained model in a model repository.

[0606] Step 4:

[0607] The terminal identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The input is conversation data from the negotiation, and the output is the identified key phrases and warnings or suggestions based on them. Specifically, the terminal uses a real-time data analysis module to convert the conversation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears.

[0608] Step 5:

[0609] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The inputs are voice tone and facial expression data, and the output is the analyzed emotional state. Specifically, the device captures the customer's facial expression with a camera, analyzes it with a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0610] Step 6:

[0611] After a sales negotiation, the user (salesperson) feeds the results back to the system. The input is the negotiation result (success or failure), and the output is an updated prediction model. Specifically, the user inputs the negotiation result into the system, and the server uses this feedback data to retrain the model, and retrains it to improve the model's accuracy.

[0612] (Application example 2)

[0613] 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."

[0614] The progress of sales negotiations and consensus building in sales activities often depend on the experience and intuition of individual sales representatives, resulting in variations in the success rate of sales negotiations. Furthermore, it is not easy to appropriately grasp and respond to the customer's emotional state, and effective methods are needed to increase the likelihood of agreement in sales negotiations. Furthermore, systems that support sales negotiations in real time lack accuracy and practicality, so the development of more advanced and effective sales negotiation support systems is necessary.

[0615] 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.

[0616] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data to extract key phrases related to success and failure, and means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases. This allows sales representatives to identify the key phrases in real time during negotiations and receive warnings or suggestions. Furthermore, by including emotion analysis means for analyzing the emotional state of the customer and means for making suggestions for adjusting the progress of the negotiations based on the emotional data analyzed by the emotion analysis means, appropriate responses can be made based on the customer's emotions, further increasing the likelihood of agreement in the negotiations. Furthermore, by including means for collecting the results of the negotiations as feedback and improving the accuracy of the model, the accuracy of the system can be continuously improved.

[0617] "Transaction data" refers to all information related to past business negotiations and negotiations, including minutes of business negotiations, transaction records, and email correspondence.

[0618] "Collection means" is a general term for the equipment and programs used to compile the necessary data and gather information.

[0619] "Analysis tools" is a general term for the equipment and algorithms used to examine collected data in detail and extract meaningful information from it.

[0620] "Key phrases" refer to important phrases or terms related to the success or failure of a deal.

[0621] "Agreement likelihood" refers to the likelihood that a particular deal will be successful and result in an agreement.

[0622] "Emotion analysis means" is a general term for devices and programs that analyze a customer's tone of voice, facial expressions, movements, etc., to identify the customer's emotional state.

[0623] "Real-time identification means" is a general term for devices and programs that perform real-time data analysis during business negotiations and instantly identify important information.

[0624] "Feedback tools" is a general term for devices and programs that systematically collect results obtained after business negotiations and are used to improve and learn from the system.

[0625] This invention is a system that uses a generation AI and an emotion engine to systematize sales negotiation agreement know-how in sales activities, and specific embodiments will be described below.

[0626] 1. Data Collection and Preprocessing

[0627] The server first collects data on past business negotiations and transactions from the company's internal database, including minutes of business negotiations, transaction records, and email correspondence, and then converts the collected data into text format for analysis.

[0628] 2. Keyphrase Extraction

[0629] The server analyzes the collected transaction data using generative AI to extract key phrases related to success and failure. This analysis uses machine learning and natural language processing (NLP) techniques. Specific tools include Hugging Face's Transformers.

[0630] 3. Creating and updating the model

[0631] The server generates and periodically updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model uses regression analysis or machine learning algorithms, such as open-source libraries like scikit-learn.

[0632] 4. Real-time support for business negotiations

[0633] The salesperson's device (e.g., smart glasses) identifies key phrases in real time during a sales call and provides warnings or suggestions as needed. The device immediately notifies the customer when key phrases such as "warranty period" or "additional services" appear during the call. Software such as OpenCV and SpeechRecognition are also used to analyze the customer's facial expressions and tone of voice and suggest appropriate responses based on their emotional state.

[0634] 5. Leveraging Emotional Engines

[0635] The device uses an emotion engine to analyze the customer's emotional state and provide feedback to the salesperson. For example, if the customer looks anxious, the device will use voice tone analysis to suggest appropriate countermeasures.

[0636] 6. Feedback on sales results

[0637] After a sales meeting is completed, the sales representative provides feedback to the system, and the server uses this feedback to further improve the accuracy of the prediction model and emotion recognition model.

[0638] Specific examples

[0639] For example, if a salesperson wearing smart glasses is asked "How long is the warranty period?" during a sales negotiation with a customer in a physical store, the system will detect the important key phrase "warranty period." If the customer looks anxious, the system will notify the salesperson to provide a more detailed explanation, thereby helping to increase the likelihood of an agreement.

[0640] Prompt Sentence Examples

[0641] When a customer asks, "How long is the product warranty period?", the system detects the important key phrase "warranty period" and notifies the sales representative. If anxiety is detected from the customer's facial expression, the system suggests a solution.

[0642] This allows salespeople to respond optimally to the needs and emotional state of the customer, thereby improving the success rate of sales negotiations.

[0643] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0644] Flow of the system program that realizes the application example

[0645] Step 1:

[0646] The server collects past transaction data from the company's internal database and converts it into text format. The collected data includes minutes of business meetings, transaction records, and email correspondence. It receives data from the company's internal database as input and outputs text data that can be analyzed.

[0647] Step 2:

[0648] The server analyzes the collected text data using generative AI and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. It receives text data as input and obtains key phrases related to success and failure as output. This processing uses Hugging Face's Transformers library.

[0649] Step 3:

[0650] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically. Specifically, it uses regression analysis and machine learning algorithms. It receives key phrases as input and obtains a predictive model as output. Machine learning libraries such as scikit-learn are used.

[0651] Step 4:

[0652] The device (e.g., smart glasses) captures audio and video in real time during a sales meeting and analyzes it. Specifically, the camera captures the customer's facial expressions and the microphone records their voice. The real-time data obtained from the camera and microphone is received as input, and the analysis results are obtained as output. OpenCV and the SpeechRecognition library are used.

[0653] Step 5:

[0654] The device uses the generative AI model from the server to identify key phrases that appear during sales negotiations in real time and provide warnings or suggestions to the sales representative as needed. It receives real-time analysis results and predictive models as input and provides notifications and suggestions as output.

[0655] Step 6:

[0656] The device uses an emotion analysis engine to analyze the voice tone and facial expression data during the negotiation to detect the customer's emotional state. It receives audio and video data as input and outputs the analysis result of the customer's emotional state.

[0657] Step 7:

[0658] The device then proposes appropriate countermeasures to the salesperson based on the results of the emotion analysis. For example, if the customer is feeling anxious, it will notify them to provide a detailed explanation. The device receives the emotion analysis results as input and proposes countermeasures as output.

[0659] Step 8:

[0660] After the negotiation is completed, the user provides feedback on the outcome (success or failure) to the system, and the server retrains the prediction model based on this data. The server receives the negotiation result data as input and obtains an improved prediction model as output.

[0661] 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.

[0662] 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.

[0663] 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.

[0664] [Third embodiment]

[0665] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0666] 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.

[0667] 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).

[0668] 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.

[0669] 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.

[0670] 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).

[0671] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[0672] 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.

[0673] 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.

[0674] 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.

[0675] 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.

[0676] 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."

[0677] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0678] 1. Data Collection

[0679] The server first collects past transaction data, including meeting minutes, transaction records, and email correspondence, from the company's internal database, extracting it and converting it into text format as needed for easier analysis.

[0680] Specific examples

[0681] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0682] 2. Keyphrase Extraction

[0683] The server analyzes the collected transaction data with generative AI to extract key phrases related to the success and failure of business negotiations. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0684] Specific examples

[0685] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0686] 3. Creating and updating the model

[0687] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0688] Specific examples

[0689] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0690] 4. Real-time support for business negotiations

[0691] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0692] Specific examples

[0693] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0694] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0695] 5. Feedback and model refinement

[0696] After the sales negotiation, the user (sales representative) provides feedback to the system, and the server uses this feedback to further refine the prediction model.

[0697] Specific examples

[0698] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0699] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0700] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0701] The processing flow will be explained below.

[0702] Step 1: Data collection

[0703] The server extracts past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and converts the data into text format for analysis.

[0704] Specific actions

[0705] The server extracts the transaction data from the database every night at midnight.

[0706] The server converts the extracted data into text format and saves it.

[0707] Step 2: Extracting Keyphrases

[0708] The server analyzes the collected transaction data and extracts key phrases associated with success and failure. This analysis is performed using generative AI, employing machine learning and natural language processing (NLP) techniques.

[0709] Specific actions

[0710] The server inputs the collected transaction data into a machine learning model.

[0711] The server uses NLP technology to identify frequently occurring key phrases from a dataset of successful and unsuccessful deals.

[0712] Step 3: Generate and update the model

[0713] The server generates and updates a model for predicting the probability of a deal being concluded based on the extracted key phrases. This model includes an algorithm for predicting whether a deal will be successful or not.

[0714] Specific actions

[0715] The server uses the extracted key phrases as input data to generate regression analysis and machine learning models to predict the likelihood of agreement.

[0716] The server periodically retrains the model with new data to improve its accuracy.

[0717] Step 4: Real-time support for business negotiations

[0718] The device identifies key phrases in real time during a sales negotiation, provides warnings and suggestions as needed, and evaluates the progress of the negotiation based on the latest key phrase list provided by the server.

[0719] Specific actions

[0720] The device analyzes speech during negotiations and notifies the sales representative when key phrases appear.

[0721] The device will display a warning if important key phrases are missing from the meeting minutes.

[0722] Step 5: Feedback on the outcome of the deal

[0723] After the sales negotiation is completed, the user provides feedback on the results of the negotiation to the system, and the server uses this feedback to further refine the prediction model.

[0724] Specific actions

[0725] The user inputs information into the system about the results of the business negotiation, such as "success" or "failure."

[0726] The server analyzes the input feedback data and retrains the model to improve its predictive accuracy.

[0727] This series of steps allows salespeople to reduce unnecessary approaches and achieve rational and efficient sales activities.

[0728] Example 1

[0729] 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."

[0730] In conventional sales activities, there was insufficient data analysis to predict the success or failure of sales negotiations, and sales relied heavily on experience. As a result, there was a need for a method to efficiently and reliably increase the likelihood of sales negotiations reaching an agreement. In addition, it was difficult to provide useful information in real time during sales negotiations, which meant that sales representatives were unable to take immediate measures.

[0731] 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.

[0732] In this invention, the server includes: means for converting past transaction data into text format using optical character recognition technology; means for analyzing transaction data using a generative AI model and inputting prompt sentences; means for analyzing collected transaction data and extracting key phrases related to success and failure; means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases; means for identifying the key phrases in real time during a sales negotiation and issuing warnings or suggestions; and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to identify important key phrases that lead to the success of a sales negotiation in real time and provide immediate countermeasures. Furthermore, the model can be updated based on feedback data to continuously improve the likelihood of agreement in sales negotiations.

[0733] "Transactional data" refers to data related to past business activities, such as minutes of business meetings, transaction records, and email correspondence.

[0734] "Optical character recognition technology" refers to the technology that converts non-text documents such as images and PDFs into text data by machine.

[0735] A "generative AI model" refers to an artificial intelligence model that uses machine learning techniques to make predictions and analyze specific input data.

[0736] A "prompt sentence" refers to a sentence in the form of input that allows a generative AI model to respond appropriately.

[0737] "Key phrases" refer to important phrases or words extracted from text data that are related to the success or failure of a business deal.

[0738] "Agreement probability" refers to the degree to which the probability of a business negotiation reaching an agreement is predicted.

[0739] A "model" refers to a computational method or algorithm constructed to make predictions or perform analysis on specific input data.

[0740] "Real-time" refers to data being processed and analyzed almost as soon as it is generated.

[0741] "Feedback" refers to entering evaluations and information about the results and process of sales negotiations into the system and reflecting them in the next forecast or analysis.

[0742] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0743] Data collection

[0744] The server first accesses the company's internal database to collect past transaction data, including meeting minutes, transaction records, and email correspondence. Because this data is often stored in PDF or image format, the server converts it into text using optical character recognition technology. For example, Tesseract OCR software is used to convert PDF files into text files, which are then stored in the database.

[0745] Keyphrase Extraction

[0746] The server inputs the collected text data into a generative AI model. The generative AI model (e.g., GPT-3) uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. In this process, only key phrases with a confidence level above a certain probability are saved as a list. An example of a prompt sentence is "Please extract important key phrases that distinguish between success and failure in business negotiation activities."

[0747] Generating and Updating Models

[0748] The server uses the extracted key phrases to generate a model that predicts the likelihood of agreement. It uses machine learning algorithms (e.g., random forests or regression analysis) to learn patterns with a high probability of success from past sales negotiation data. The model is periodically updated with new data. For example, it performs regression analysis based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0749] Real-time support for business negotiations

[0750] The terminal (salesperson's device) monitors the sales negotiation record in real time during the negotiation and identifies key phrases. Based on the key phrase list provided by the server, it gives warnings and suggestions to the salesperson depending on the progress of the negotiation. For example, when the key phrase "customer support" is detected, the terminal displays a pop-up notification and makes a suggestion such as "This key phrase will increase the probability of agreement."

[0751] Feedback and model refinement

[0752] After a sales meeting is completed, the user (sales representative) provides feedback to the system on the results. They enter whether the meeting was a success or failure, and record the details of each case. The server uses this feedback to retrain the prediction model and improve prediction accuracy. For example, if a user enters feedback such as "This meeting was successful" after the meeting is completed, the server analyzes that data and the list of key phrases detected during the meeting to retrain the prediction model.

[0753] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0754] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0755] Step 1:

[0756] The server accesses the company's internal database and collects past transaction data. The transaction data includes meeting minutes, transaction records, and email correspondence. The server extracts this data from the database using SQL queries. For example, the server executes the following query: SELECT FROM meeting data WHERE date BETWEEN '2022-01-01' AND '2023-01-01'. The extracted data is converted to text format. Meeting minutes in PDF format are converted to text using OCR technology. The input is a PDF file, and the output is a text file.

[0757] Specific behavior:

[0758] The server connects to the database and extracts the minutes of business meetings from the past year using an SQL query.

[0759] After extraction, the PDF files are converted to text files using Tesseract OCR software.

[0760] The converted text file is formatted and saved in a unified format.

[0761] Step 2:

[0762] The server inputs the collected text data into a generative AI model (e.g., GPT-3). The generative AI model uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. The input prompt is "Please extract important key phrases that distinguish between success and failure in business negotiation activities," and the model outputs key phrases based on this. The input is text data, and the output is a list of key phrases.

[0763] Specific behavior:

[0764] The server inputs text data into the generative AI model.

[0765] Send the prompt "Please extract the key phrases that distinguish between success and failure in sales negotiation activities" to the model.

[0766] The generative AI model identifies key phrases and outputs them in a list format.

[0767] The server stores the extracted key phrases in a database.

[0768] Step 3:

[0769] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This process uses a machine learning algorithm (e.g., random forest or regression analysis). The input is the list of key phrases and past sales performance data, and the output is the predictive model.

[0770] Specific behavior:

[0771] The server performs regression analysis using sales negotiation data with a success rate of 80% or higher.

[0772] It learns the relationship between the frequency of key phrases and the success rate of sales negotiations, and generates a model that predicts the probability of agreement.

[0773] The trained model is stored on the server and periodically updated with new data.

[0774] Step 4:

[0775] The terminal monitors data entered in real time during sales negotiations and identifies key phrases. Based on the key phrase list provided by the server, it issues warnings or suggestions to sales representatives according to the progress of the negotiations. The input is real-time sales negotiation records, and the output is warning or suggestion notifications.

[0776] Specific behavior:

[0777] The device monitors conversations and text input during business negotiations in real time.

[0778] If the key phrase "customer support" is identified, the device will display a pop-up notification.

[0779] If an important keyphrase is missing, the terminal will display a warning message.

[0780] Step 5:

[0781] After a sales negotiation is completed, the user provides feedback to the system on the results, entering whether the negotiation was successful or not and adding detailed comments. The server uses this feedback data to retrain the prediction model and improve its accuracy. The input is the feedback data, and the output is an updated prediction model.

[0782] Specific behavior:

[0783] After the negotiation is completed, the user inputs feedback on the success of the negotiation into the system.

[0784] The server analyzes the feedback data and the key phrase data detected during the business negotiation.

[0785] Retrain and update the predictive model.

[0786] The updated model will be applied to the next deal.

[0787] In this way, this system provides support to improve the likelihood of agreement in business negotiations through each step.

[0788] (Application example 1)

[0789] 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."

[0790] Conventional sales negotiation support systems for sales activities are limited to methods that utilize past transaction data to predict the likelihood of success of negotiations, making them difficult to apply to other business environments. Furthermore, there are limitations to efficient product picking methods at logistics facilities, creating a need for improved work efficiency. Currently, there is a lack of means to make specific suggestions for optimizing work efficiency in real time during picking work. This means that optimal support for workers and robots to work efficiently cannot be provided.

[0791] 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.

[0792] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting agreement probability based on the extracted key phrases, means for identifying the key phrases in real time during negotiations and issuing warnings or suggestions, means for collecting the results of the negotiations as feedback and improving the accuracy of the model, and means for applying the model to support operational efficiency in logistics facilities and making suggestions in real time when a robot performs product picking work. This makes it possible to optimize the efficiency of picking work in logistics facilities in addition to sales activities in real time.

[0793] "Transaction data" refers to data related to sales activities, including minutes of business negotiations, transaction records, email exchanges, etc.

[0794] "Key phrases" are specific important words or phrases related to the success or failure of a business deal.

[0795] "Agreement probability" is an indicator that indicates the probability that a business negotiation will be successful.

[0796] A "model" is a computational method or algorithm used to analyze collected data and predict the likelihood of agreement.

[0797] A "logistics facility" is a facility for storing goods and for picking and shipping operations.

[0798] "Product picking work" refers to the work of picking out products at a logistics facility, packaging them, and shipping them.

[0799] A "robot" is a mechanical device that automates and efficiently performs product picking tasks at logistics facilities.

[0800] "Real-time" is a concept that refers to immediate response during negotiations or product picking.

[0801] A "proposal" is an act of giving advice or instructions to efficiently advance business negotiations or picking work.

[0802] "Feedback" refers to reflecting the results of business negotiations and picking operations in the system.

[0803] "Model accuracy" is an indicator of how closely the results predicted by the model match the actual results.

[0804] The present invention relates to a system for improving the efficiency of product picking operations in sales activities and logistics facilities. Specific embodiments of the system will be described below.

[0805] 1. Data Collection

[0806] The server first collects past transaction data. This data includes minutes of business meetings, transaction records, email correspondence, and more. It also collects past picking data from logistics facilities. The server extracts this data from the company's internal database and converts it into text format as needed, making it easier to analyze.

[0807] 2. Keyphrase Extraction

[0808] The server uses generative AI to analyze the collected transaction data and extract key phrases related to the success or failure of sales negotiations. Similarly, it extracts key phrases related to efficient work from the picking data. This analysis uses machine learning and natural language processing (NLP) techniques. Specifically, it uses TfidfVectorizer and RandomForestClassifier.

[0809] 3. Creating and updating the model

[0810] The server generates and updates a model that predicts the likelihood of a negotiation agreement and the efficiency of picking work based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of negotiations and picking work.

[0811] 4. Real-time support for sales negotiations and picking

[0812] The terminal identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as necessary. It evaluates the progress of negotiations and picking work based on a key phrase list provided by the server. For example, if the key phrase "ROI (return on investment)" appears during negotiations, the sales representative is notified that the likelihood of agreement is increasing. Also, if the key phrase "high-demand product" appears during picking work, an efficient picking order is suggested based on that.

[0813] 5. Feedback and model refinement

[0814] After a sales negotiation or picking task is completed, the user provides feedback to the system. The server uses this feedback to further refine the prediction model. For example, the user can input feedback such as "success" or "failure" as the outcome of the negotiation. The server analyzes this feedback data and retrains the model to improve its prediction accuracy.

[0815] As a specific example, the system identifies key phrases such as "high-demand products" and "high-priority items" and suggests an efficient picking order based on them.

[0816] Example prompts to input to a generative AI model:

[0817] "Please analyze past picking work logs, extract key phrases to improve picking efficiency, and suggest the optimal picking order in real time."

[0818] This system makes it possible to optimize the efficiency of sales activities as well as picking operations at logistics facilities in real time.

[0819] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0820] Step 1:

[0821] The server collects past transaction data and picking data. It extracts data from the transaction database and picking database as input and obtains data converted into text format as output. Specifically, it converts data such as minutes, transaction records, email correspondence, and picking logs into a format that is easy to centrally manage.

[0822] Step 2:

[0823] The server analyzes the collected transaction and picking data to extract key phrases related to success and failure. It uses the collected text data as input and generates a list of key phrases as output. Specifically, it vectorizes the text data using TfidfVectorizer and identifies important key phrases using RandomForestClassifier.

[0824] Step 3:

[0825] The server generates and updates models that predict the likelihood of agreement in negotiations and the efficiency of picking work based on the extracted key phrases. It uses the key phrase list as input and generates a predictive model as output. Specifically, it analyzes the frequency and relationships of key phrases and builds the model using regression analysis and machine learning algorithms.

[0826] Step 4:

[0827] The device identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as needed. It uses real-time negotiation voice data or picking logs as input and generates warning or suggestion messages as output. Specifically, it works in conjunction with voice recognition software (e.g., Google Speech-to-Text) to detect pre-defined key phrases in real time.

[0828] Step 5:

[0829] After a sales negotiation or picking task is completed, the user provides feedback to the system. The results of the negotiation (success or failure) or the results of the picking task (achievement rate, number of errors) are input to the system, and feedback data is generated as output. Specifically, the results are entered through a mobile application or web interface and sent to the server as feedback data.

[0830] Step 6:

[0831] The server retrains the predictive model based on the feedback to improve its accuracy. It uses the feedback data as input and generates an improved predictive model as output. Specifically, it feeds the newly collected feedback data into a machine learning algorithm to optimize the model's parameters.

[0832] 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.

[0833] The present invention relates to a system that uses a generation AI and an emotion engine to systematize know-how for reaching agreements in sales negotiations, and specific embodiments thereof are described below.

[0834] 1. Data Collection

[0835] The server first collects past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and then converts the data into text format for analysis.

[0836] Specific examples

[0837] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[0838] 2. Keyphrase Extraction

[0839] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[0840] Specific examples

[0841] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[0842] 3. Creating and updating the model

[0843] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[0844] Specific examples

[0845] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[0846] 4. Real-time support for business negotiations

[0847] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[0848] Specific examples

[0849] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[0850] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[0851] 5. Leveraging Emotional Engines

[0852] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0853] Specific examples

[0854] The device uses a camera to analyze the customer's facial expressions during sales negotiations and notifies the sales representative if emotions such as anxiety or excitement are detected.

[0855] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction and makes suggestions to adjust the progress of the business negotiations.

[0856] 6. Feedback on sales results

[0857] After the sales negotiation, the user (salesperson) provides feedback to the system, and the server uses this feedback to further refine the prediction model and emotion recognition model.

[0858] Specific examples

[0859] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[0860] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[0861] This system allows salespeople to grasp important key phrases and the emotional state of the customer during the first sales meeting, increasing the likelihood of an agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0862] The processing flow will be explained below.

[0863] Step 1: Data collection

[0864] The server extracts past transaction data from the company's internal database, including business meeting minutes, transaction records, and email correspondence. The collected data is then converted into text format, making it suitable for analysis.

[0865] Specific actions

[0866] The server automatically extracts transaction data from the database every night.

[0867] The server converts the extracted data into text format and prepares it into an analyzable file format.

[0868] Step 2: Extracting Keyphrases

[0869] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis utilizes machine learning and natural language processing (NLP) techniques.

[0870] Specific actions

[0871] The server inputs the collected transaction data into a machine learning model and performs NLP analysis.

[0872] The server compares data from successful and unsuccessful transactions to identify frequently occurring key phrases.

[0873] Step 3: Generate and update the model

[0874] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically.

[0875] Specific actions

[0876] The server creates a predictive model using regression analysis and machine learning algorithms based on key phrases that frequently appear in business negotiations with a high success rate.

[0877] The server periodically retrains the model with newly collected data to improve prediction accuracy.

[0878] Step 4: Real-time support for business negotiations

[0879] The terminal, as a device for sales representatives, identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the latest key phrase list provided by the server.

[0880] Specific actions

[0881] The device uses voice recognition technology to analyze speech during negotiations and notify the sales representative when important key phrases appear.

[0882] The terminal displays a warning if an important key phrase is missing when entering business meeting minutes.

[0883] Step 5: Leverage the Emotion Engine

[0884] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[0885] Specific actions

[0886] The device uses a camera to analyze the customer's facial expressions during the sales negotiation and recognizes their emotional state (e.g., anxiety, excitement, satisfaction, etc.) in real time.

[0887] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction, providing the sales representative with information to adjust the progress of the negotiation and the content of the proposal.

[0888] Step 6: Feedback on the outcome of the deal

[0889] After a sales meeting, the user (sales representative) provides feedback to the system, and the server uses this feedback to improve the accuracy of the prediction model and emotion engine.

[0890] Specific actions

[0891] The user enters the outcome of the negotiation (success or failure) and the reason for it into the system.

[0892] The server analyzes the feedback data and retrains the predictive model and emotion engine to improve accuracy.

[0893] Example 2

[0894] 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."

[0895] To increase the likelihood of agreement in sales negotiations, it is important to utilize past transaction data to identify factors related to success and failure. However, conventional methods make it difficult to efficiently analyze these factors and support sales negotiations in real time. In addition, more effective sales negotiation support is required by understanding the customer's emotional state, but there is a lack of technology to perform emotional analysis in real time. To solve these issues, a comprehensive and efficient system is required.

[0896] 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.

[0897] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases, means for identifying key phrases in real time during a sales negotiation and issuing a warning or suggestion, means for analyzing the emotional state of the customer in real time using a sentiment analysis engine and notifying the sales representative, and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to accurately grasp important factors in sales negotiations and increase the likelihood of agreement in sales negotiations.

[0898] "Transaction data" refers to a series of records related to business negotiations, such as minutes of past business negotiations, transaction records, and email exchanges.

[0899] "Key phrases" refer to important words or short phrases related to the success or failure of a business deal, extracted through analysis of transaction data.

[0900] An "agreement probability prediction model" refers to a mathematical model that includes calculation methods and algorithms for predicting the success or failure of a business negotiation based on extracted key phrases.

[0901] "Generative AI" refers to artificial intelligence that uses machine learning and natural language processing techniques to analyze collected transaction data and extract key phrases.

[0902] "Feedback" refers to the information on the results of sales negotiations that sales representatives enter into the system after the negotiations are completed, and is data used to improve the accuracy of the model.

[0903] An "emotion analysis engine" refers to technology that analyzes data such as a customer's tone of voice and facial expressions to recognize their emotional state in real time.

[0904] "Real-time analysis" refers to a processing method that processes data instantly while a sales negotiation is in progress and provides sales representatives with the necessary information and warnings immediately.

[0905] "Salesperson's device" refers to a device (e.g., PC, tablet, smartphone) that receives real-time information and displays alerts and suggestions during a sales meeting.

[0906] The present invention relates to a system for systematizing sales negotiation know-how using a generation AI and a sentiment analysis engine. Specific embodiments of the system are described below.

[0907] First, the server collects past transaction data from the company's internal database. This transaction data includes sales meeting minutes, transaction records, and email correspondence, and converts it into text format and saves it. For example, the server uses an SQL query to extract sales data from January 2022 to January 2023, and then uses a Python script to save it as a text JSON file.

[0908] Next, the server analyzes the collected transaction data using a generative AI (e.g., GPT-4) to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques. Specifically, the server preprocesses the collected text data using a natural language processing library (e.g., spaCy, NLTK), inputs it into the generative AI to extract key phrases, and stores the extracted key phrases in a database.

[0909] The server then generates and updates a model to predict the likelihood of agreement based on the extracted key phrases. This model includes algorithms (e.g., regression analysis, machine learning) for predicting the success or failure of a deal. Specifically, the server combines the extracted key phrases with the deal outcome data, trains a regression model or classifier using a machine learning library (e.g., scikit-learn), and stores the trained model in a model repository on the server.

[0910] The terminal (sales representative's device) identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The progress of the negotiation is evaluated based on the key phrase list provided by the server. Specifically, the terminal uses a real-time data analysis module to convert the conversation during the negotiation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears. It also issues a warning if an important key phrase is omitted when creating minutes of the negotiation.

[0911] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The emotion analysis engine recognizes emotions by analyzing data points such as the customer's voice tone and facial expressions. Specifically, the device captures the customer's facial expressions with a camera, performs image analysis using a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0912] Finally, after the sales negotiation, the user (salesperson) provides feedback to the system on the outcome. The server uses this feedback to further refine the prediction model and emotion recognition model. Specifically, the user enters "success" or "failure" as the outcome of the negotiation into the system, and the server uses this feedback data as a new learning dataset to retrain the model.

[0913] Example prompt

[0914] 1. "Extract key phrases related to success and failure from past transaction data."

[0915] 2. "Suggest a way to analyze a customer's tone of voice and facial expressions to signal their emotional state during a sales conversation."

[0916] 3. "How can we refine our predictive models using data from sales rep feedback on deal outcomes?"

[0917] By utilizing this system, sales representatives can accurately grasp the important elements of sales negotiations, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[0918] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0919] Step 1:

[0920] The server collects past transaction data from the company's internal database. Input data includes meeting minutes, transaction records, and email correspondence. This data is converted into text format and output. Specifically, the server uses SQL queries to extract the transaction data, converts it into a text-format JSON file using a Python script, and saves it.

[0921] Step 2:

[0922] The server analyzes the collected transaction data using a generation AI and extracts key phrases related to success and failure. The input is text-format transaction data, and the output is the extracted key phrases. Specifically, the server performs preprocessing using a natural language processing library (e.g., spaCy, NLTK), inputs the data into the generation AI to extract key phrases, and stores them in a database.

[0923] Step 3:

[0924] The server generates and updates a model that predicts the probability of agreement based on the extracted key phrases. The inputs are key phrases and negotiation outcome data, and the output is a predictive model. Specifically, the server creates a training dataset using a machine learning library (e.g., scikit-learn), trains a regression model or classifier, and saves the trained model in a model repository.

[0925] Step 4:

[0926] The terminal identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The input is conversation data from the negotiation, and the output is the identified key phrases and warnings or suggestions based on them. Specifically, the terminal uses a real-time data analysis module to convert the conversation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears.

[0927] Step 5:

[0928] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The inputs are voice tone and facial expression data, and the output is the analyzed emotional state. Specifically, the device captures the customer's facial expression with a camera, analyzes it with a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[0929] Step 6:

[0930] After a sales negotiation, the user (salesperson) feeds the results back to the system. The input is the negotiation result (success or failure), and the output is an updated prediction model. Specifically, the user inputs the negotiation result into the system, and the server uses this feedback data to retrain the model, and retrains it to improve the model's accuracy.

[0931] (Application example 2)

[0932] 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."

[0933] The progress of sales negotiations and consensus building in sales activities often depend on the experience and intuition of individual sales representatives, resulting in variations in the success rate of sales negotiations. Furthermore, it is not easy to appropriately grasp and respond to the customer's emotional state, and effective methods are needed to increase the likelihood of agreement in sales negotiations. Furthermore, systems that support sales negotiations in real time lack accuracy and practicality, so the development of more advanced and effective sales negotiation support systems is necessary.

[0934] 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.

[0935] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data to extract key phrases related to success and failure, and means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases. This allows sales representatives to identify the key phrases in real time during negotiations and receive warnings or suggestions. Furthermore, by including emotion analysis means for analyzing the emotional state of the customer and means for making suggestions for adjusting the progress of the negotiations based on the emotional data analyzed by the emotion analysis means, appropriate responses can be made based on the customer's emotions, further increasing the likelihood of agreement in the negotiations. Furthermore, by including means for collecting the results of the negotiations as feedback and improving the accuracy of the model, the accuracy of the system can be continuously improved.

[0936] "Transaction data" refers to all information related to past business negotiations and negotiations, including minutes of business negotiations, transaction records, and email correspondence.

[0937] "Collection means" is a general term for the equipment and programs used to compile the necessary data and gather information.

[0938] "Analysis tools" is a general term for the equipment and algorithms used to examine collected data in detail and extract meaningful information from it.

[0939] "Key phrases" refer to important phrases or terms related to the success or failure of a deal.

[0940] "Agreement likelihood" refers to the likelihood that a particular deal will be successful and result in an agreement.

[0941] "Emotion analysis means" is a general term for devices and programs that analyze a customer's tone of voice, facial expressions, movements, etc., to identify the customer's emotional state.

[0942] "Real-time identification means" is a general term for devices and programs that perform real-time data analysis during business negotiations and instantly identify important information.

[0943] "Feedback tools" is a general term for devices and programs that systematically collect results obtained after business negotiations and are used to improve and learn from the system.

[0944] This invention is a system that uses a generation AI and an emotion engine to systematize sales negotiation agreement know-how in sales activities, and specific embodiments will be described below.

[0945] 1. Data Collection and Preprocessing

[0946] The server first collects data on past business negotiations and transactions from the company's internal database, including minutes of business negotiations, transaction records, and email correspondence, and then converts the collected data into text format for analysis.

[0947] 2. Keyphrase Extraction

[0948] The server analyzes the collected transaction data using generative AI to extract key phrases related to success and failure. This analysis uses machine learning and natural language processing (NLP) techniques. Specific tools include Hugging Face's Transformers.

[0949] 3. Creating and updating the model

[0950] The server generates and periodically updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model uses regression analysis or machine learning algorithms, such as open-source libraries like scikit-learn.

[0951] 4. Real-time support for business negotiations

[0952] The salesperson's device (e.g., smart glasses) identifies key phrases in real time during a sales call and provides warnings or suggestions as needed. The device immediately notifies the customer when key phrases such as "warranty period" or "additional services" appear during the call. Software such as OpenCV and SpeechRecognition are also used to analyze the customer's facial expressions and tone of voice and suggest appropriate responses based on their emotional state.

[0953] 5. Leveraging Emotional Engines

[0954] The device uses an emotion engine to analyze the customer's emotional state and provide feedback to the salesperson. For example, if the customer looks anxious, the device will use voice tone analysis to suggest appropriate countermeasures.

[0955] 6. Feedback on sales results

[0956] After a sales meeting is completed, the sales representative provides feedback to the system, and the server uses this feedback to further improve the accuracy of the prediction model and emotion recognition model.

[0957] Specific examples

[0958] For example, if a salesperson wearing smart glasses is asked "How long is the warranty period?" during a sales negotiation with a customer in a physical store, the system will detect the important key phrase "warranty period." If the customer looks anxious, the system will notify the salesperson to provide a more detailed explanation, thereby helping to increase the likelihood of an agreement.

[0959] Prompt Sentence Examples

[0960] When a customer asks, "How long is the product warranty period?", the system detects the important key phrase "warranty period" and notifies the sales representative. If anxiety is detected from the customer's facial expression, the system suggests a solution.

[0961] This allows salespeople to respond optimally to the needs and emotional state of the customer, thereby improving the success rate of sales negotiations.

[0962] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0963] Flow of the system program that realizes the application example

[0964] Step 1:

[0965] The server collects past transaction data from the company's internal database and converts it into text format. The collected data includes minutes of business meetings, transaction records, and email correspondence. It receives data from the company's internal database as input and outputs text data that can be analyzed.

[0966] Step 2:

[0967] The server analyzes the collected text data using generative AI and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. It receives text data as input and obtains key phrases related to success and failure as output. This processing uses Hugging Face's Transformers library.

[0968] Step 3:

[0969] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically. Specifically, it uses regression analysis and machine learning algorithms. It receives key phrases as input and obtains a predictive model as output. Machine learning libraries such as scikit-learn are used.

[0970] Step 4:

[0971] The device (e.g., smart glasses) captures audio and video in real time during a sales meeting and analyzes it. Specifically, the camera captures the customer's facial expressions and the microphone records their voice. The real-time data obtained from the camera and microphone is received as input, and the analysis results are obtained as output. OpenCV and the SpeechRecognition library are used.

[0972] Step 5:

[0973] The device uses the generative AI model from the server to identify key phrases that appear during sales negotiations in real time and provide warnings or suggestions to the sales representative as needed. It receives real-time analysis results and predictive models as input and provides notifications and suggestions as output.

[0974] Step 6:

[0975] The device uses an emotion analysis engine to analyze the voice tone and facial expression data during the negotiation to detect the customer's emotional state. It receives audio and video data as input and outputs the analysis result of the customer's emotional state.

[0976] Step 7:

[0977] The device then proposes appropriate countermeasures to the salesperson based on the results of the emotion analysis. For example, if the customer is feeling anxious, it will notify them to provide a detailed explanation. The device receives the emotion analysis results as input and proposes countermeasures as output.

[0978] Step 8:

[0979] After the negotiation is completed, the user provides feedback on the outcome (success or failure) to the system, and the server retrains the prediction model based on this data. The server receives the negotiation result data as input and obtains an improved prediction model as output.

[0980] 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.

[0981] 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.

[0982] 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.

[0983] [Fourth embodiment]

[0984] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0985] 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.

[0986] 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).

[0987] 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.

[0988] 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.

[0989] 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).

[0990] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for 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.

[0991] 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.

[0992] 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.

[0993] 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.

[0994] 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.

[0995] 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.

[0996] 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."

[0997] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[0998] 1. Data Collection

[0999] The server first collects past transaction data, including meeting minutes, transaction records, and email correspondence, from the company's internal database, extracting it and converting it into text format as needed for easier analysis.

[1000] Specific examples

[1001] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[1002] 2. Keyphrase Extraction

[1003] The server analyzes the collected transaction data with generative AI to extract key phrases related to the success and failure of business negotiations. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[1004] Specific examples

[1005] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[1006] 3. Creating and updating the model

[1007] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[1008] Specific examples

[1009] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[1010] 4. Real-time support for business negotiations

[1011] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[1012] Specific examples

[1013] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[1014] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[1015] 5. Feedback and model refinement

[1016] After the sales negotiation, the user (sales representative) provides feedback to the system, and the server uses this feedback to further refine the prediction model.

[1017] Specific examples

[1018] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[1019] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[1020] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[1021] The processing flow will be explained below.

[1022] Step 1: Data collection

[1023] The server extracts past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and converts the data into text format for analysis.

[1024] Specific actions

[1025] The server extracts the transaction data from the database every night at midnight.

[1026] The server converts the extracted data into text format and saves it.

[1027] Step 2: Extracting Keyphrases

[1028] The server analyzes the collected transaction data and extracts key phrases associated with success and failure. This analysis is performed using generative AI, employing machine learning and natural language processing (NLP) techniques.

[1029] Specific actions

[1030] The server inputs the collected transaction data into a machine learning model.

[1031] The server uses NLP technology to identify frequently occurring key phrases from a dataset of successful and unsuccessful deals.

[1032] Step 3: Generate and update the model

[1033] The server generates and updates a model for predicting the probability of a deal being concluded based on the extracted key phrases. This model includes an algorithm for predicting whether a deal will be successful or not.

[1034] Specific actions

[1035] The server uses the extracted key phrases as input data to generate regression analysis and machine learning models to predict the likelihood of agreement.

[1036] The server periodically retrains the model with new data to improve its accuracy.

[1037] Step 4: Real-time support for business negotiations

[1038] The device identifies key phrases in real time during a sales negotiation, provides warnings and suggestions as needed, and evaluates the progress of the negotiation based on the latest key phrase list provided by the server.

[1039] Specific actions

[1040] The device analyzes speech during negotiations and notifies the sales representative when key phrases appear.

[1041] The device will display a warning if important key phrases are missing from the meeting minutes.

[1042] Step 5: Feedback on the outcome of the deal

[1043] After the sales negotiation is completed, the user provides feedback on the results of the negotiation to the system, and the server uses this feedback to further refine the prediction model.

[1044] Specific actions

[1045] The user inputs information into the system about the results of the business negotiation, such as "success" or "failure."

[1046] The server analyzes the input feedback data and retrains the model to improve its predictive accuracy.

[1047] This series of steps allows salespeople to reduce unnecessary approaches and achieve rational and efficient sales activities.

[1048] Example 1

[1049] 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."

[1050] In conventional sales activities, there was insufficient data analysis to predict the success or failure of sales negotiations, and sales relied heavily on experience. As a result, there was a need for a method to efficiently and reliably increase the likelihood of sales negotiations reaching an agreement. In addition, it was difficult to provide useful information in real time during sales negotiations, which meant that sales representatives were unable to take immediate measures.

[1051] 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.

[1052] In this invention, the server includes: means for converting past transaction data into text format using optical character recognition technology; means for analyzing transaction data using a generative AI model and inputting prompt sentences; means for analyzing collected transaction data and extracting key phrases related to success and failure; means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases; means for identifying the key phrases in real time during a sales negotiation and issuing warnings or suggestions; and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to identify important key phrases that lead to the success of a sales negotiation in real time and provide immediate countermeasures. Furthermore, the model can be updated based on feedback data to continuously improve the likelihood of agreement in sales negotiations.

[1053] "Transactional data" refers to data related to past business activities, such as minutes of business meetings, transaction records, and email correspondence.

[1054] "Optical character recognition technology" refers to the technology that converts non-text documents such as images and PDFs into text data by machine.

[1055] A "generative AI model" refers to an artificial intelligence model that uses machine learning techniques to make predictions and analyze specific input data.

[1056] A "prompt sentence" refers to a sentence in the form of input that allows a generative AI model to respond appropriately.

[1057] "Key phrases" refer to important phrases or words extracted from text data that are related to the success or failure of a business deal.

[1058] "Agreement probability" refers to the degree to which the probability of a business negotiation reaching an agreement is predicted.

[1059] A "model" refers to a computational method or algorithm constructed to make predictions or perform analysis on specific input data.

[1060] "Real-time" refers to data being processed and analyzed almost as soon as it is generated.

[1061] "Feedback" refers to entering evaluations and information about the results and process of sales negotiations into the system and reflecting them in the next forecast or analysis.

[1062] The present invention relates to a system that uses generation AI to systematize know-how for business negotiations in sales activities, and specific embodiments thereof will be described below.

[1063] Data collection

[1064] The server first accesses the company's internal database to collect past transaction data, including meeting minutes, transaction records, and email correspondence. Because this data is often stored in PDF or image format, the server converts it into text using optical character recognition technology. For example, Tesseract OCR software is used to convert PDF files into text files, which are then stored in the database.

[1065] Keyphrase Extraction

[1066] The server inputs the collected text data into a generative AI model. The generative AI model (e.g., GPT-3) uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. In this process, only key phrases with a confidence level above a certain probability are saved as a list. An example of a prompt sentence is "Please extract important key phrases that distinguish between success and failure in business negotiation activities."

[1067] Generating and Updating Models

[1068] The server uses the extracted key phrases to generate a model that predicts the likelihood of agreement. It uses machine learning algorithms (e.g., random forests or regression analysis) to learn patterns with a high probability of success from past sales negotiation data. The model is periodically updated with new data. For example, it performs regression analysis based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[1069] Real-time support for business negotiations

[1070] The terminal (salesperson's device) monitors the sales negotiation record in real time during the negotiation and identifies key phrases. Based on the key phrase list provided by the server, it gives warnings and suggestions to the salesperson depending on the progress of the negotiation. For example, when the key phrase "customer support" is detected, the terminal displays a pop-up notification and makes a suggestion such as "This key phrase will increase the probability of agreement."

[1071] Feedback and model refinement

[1072] After a sales meeting is completed, the user (sales representative) provides feedback to the system on the results. They enter whether the meeting was a success or failure, and record the details of each case. The server uses this feedback to retrain the prediction model and improve prediction accuracy. For example, if a user enters feedback such as "This meeting was successful" after the meeting is completed, the server analyzes that data and the list of key phrases detected during the meeting to retrain the prediction model.

[1073] This system allows sales representatives to grasp important key phrases during the first sales meeting, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[1074] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1075] Step 1:

[1076] The server accesses the company's internal database and collects past transaction data. The transaction data includes meeting minutes, transaction records, and email correspondence. The server extracts this data from the database using SQL queries. For example, the server executes the following query: SELECT FROM meeting data WHERE date BETWEEN '2022-01-01' AND '2023-01-01'. The extracted data is converted to text format. Meeting minutes in PDF format are converted to text using OCR technology. The input is a PDF file, and the output is a text file.

[1077] Specific behavior:

[1078] The server connects to the database and extracts the minutes of business meetings from the past year using an SQL query.

[1079] After extraction, the PDF files are converted to text files using Tesseract OCR software.

[1080] The converted text file is formatted and saved in a unified format.

[1081] Step 2:

[1082] The server inputs the collected text data into a generative AI model (e.g., GPT-3). The generative AI model uses machine learning and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. The input prompt is "Please extract important key phrases that distinguish between success and failure in business negotiation activities," and the model outputs key phrases based on this. The input is text data, and the output is a list of key phrases.

[1083] Specific behavior:

[1084] The server inputs text data into the generative AI model.

[1085] Send the prompt "Please extract the key phrases that distinguish between success and failure in sales negotiation activities" to the model.

[1086] The generative AI model identifies key phrases and outputs them in a list format.

[1087] The server stores the extracted key phrases in a database.

[1088] Step 3:

[1089] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This process uses a machine learning algorithm (e.g., random forest or regression analysis). The input is the list of key phrases and past sales performance data, and the output is the predictive model.

[1090] Specific behavior:

[1091] The server performs regression analysis using sales negotiation data with a success rate of 80% or higher.

[1092] It learns the relationship between the frequency of key phrases and the success rate of sales negotiations, and generates a model that predicts the probability of agreement.

[1093] The trained model is stored on the server and periodically updated with new data.

[1094] Step 4:

[1095] The terminal monitors data entered in real time during sales negotiations and identifies key phrases. Based on the key phrase list provided by the server, it issues warnings or suggestions to sales representatives according to the progress of the negotiations. The input is real-time sales negotiation records, and the output is warning or suggestion notifications.

[1096] Specific behavior:

[1097] The device monitors conversations and text input during business negotiations in real time.

[1098] If the key phrase "customer support" is identified, the device will display a pop-up notification.

[1099] If an important keyphrase is missing, the terminal will display a warning message.

[1100] Step 5:

[1101] After a sales negotiation is completed, the user provides feedback to the system on the results, entering whether the negotiation was successful or not and adding detailed comments. The server uses this feedback data to retrain the prediction model and improve its accuracy. The input is the feedback data, and the output is an updated prediction model.

[1102] Specific behavior:

[1103] After the negotiation is completed, the user inputs feedback on the success of the negotiation into the system.

[1104] The server analyzes the feedback data and the key phrase data detected during the business negotiation.

[1105] Retrain and update the predictive model.

[1106] The updated model will be applied to the next deal.

[1107] In this way, this system provides support to improve the likelihood of agreement in business negotiations through each step.

[1108] (Application example 1)

[1109] 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."

[1110] Conventional sales negotiation support systems for sales activities are limited to methods that utilize past transaction data to predict the likelihood of success of negotiations, making them difficult to apply to other business environments. Furthermore, there are limitations to efficient product picking methods at logistics facilities, creating a need for improved work efficiency. Currently, there is a lack of means to make specific suggestions for optimizing work efficiency in real time during picking work. This means that optimal support for workers and robots to work efficiently cannot be provided.

[1111] 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.

[1112] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting agreement probability based on the extracted key phrases, means for identifying the key phrases in real time during negotiations and issuing warnings or suggestions, means for collecting the results of the negotiations as feedback and improving the accuracy of the model, and means for applying the model to support operational efficiency in logistics facilities and making suggestions in real time when a robot performs product picking work. This makes it possible to optimize the efficiency of picking work in logistics facilities in addition to sales activities in real time.

[1113] "Transaction data" refers to data related to sales activities, including minutes of business negotiations, transaction records, email exchanges, etc.

[1114] "Key phrases" are specific important words or phrases related to the success or failure of a business deal.

[1115] "Agreement probability" is an indicator that indicates the probability that a business negotiation will be successful.

[1116] A "model" is a computational method or algorithm used to analyze collected data and predict the likelihood of agreement.

[1117] A "logistics facility" is a facility for storing goods and for picking and shipping operations.

[1118] "Product picking work" refers to the work of picking out products at a logistics facility, packaging them, and shipping them.

[1119] A "robot" is a mechanical device that automates and efficiently performs product picking tasks at logistics facilities.

[1120] "Real-time" is a concept that refers to immediate response during negotiations or product picking.

[1121] A "proposal" is an act of giving advice or instructions to efficiently advance business negotiations or picking work.

[1122] "Feedback" refers to reflecting the results of business negotiations and picking operations in the system.

[1123] "Model accuracy" is an indicator of how closely the results predicted by the model match the actual results.

[1124] The present invention relates to a system for improving the efficiency of product picking operations in sales activities and logistics facilities. Specific embodiments of the system will be described below.

[1125] 1. Data Collection

[1126] The server first collects past transaction data. This data includes minutes of business meetings, transaction records, email correspondence, and more. It also collects past picking data from logistics facilities. The server extracts this data from the company's internal database and converts it into text format as needed, making it easier to analyze.

[1127] 2. Keyphrase Extraction

[1128] The server uses generative AI to analyze the collected transaction data and extract key phrases related to the success or failure of sales negotiations. Similarly, it extracts key phrases related to efficient work from the picking data. This analysis uses machine learning and natural language processing (NLP) techniques. Specifically, it uses TfidfVectorizer and RandomForestClassifier.

[1129] 3. Creating and updating the model

[1130] The server generates and updates a model that predicts the likelihood of a negotiation agreement and the efficiency of picking work based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of negotiations and picking work.

[1131] 4. Real-time support for sales negotiations and picking

[1132] The terminal identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as necessary. It evaluates the progress of negotiations and picking work based on a key phrase list provided by the server. For example, if the key phrase "ROI (return on investment)" appears during negotiations, the sales representative is notified that the likelihood of agreement is increasing. Also, if the key phrase "high-demand product" appears during picking work, an efficient picking order is suggested based on that.

[1133] 5. Feedback and model refinement

[1134] After a sales negotiation or picking task is completed, the user provides feedback to the system. The server uses this feedback to further refine the prediction model. For example, the user can input feedback such as "success" or "failure" as the outcome of the negotiation. The server analyzes this feedback data and retrains the model to improve its prediction accuracy.

[1135] As a specific example, the system identifies key phrases such as "high-demand products" and "high-priority items" and suggests an efficient picking order based on them.

[1136] Example prompts to input to a generative AI model:

[1137] "Please analyze past picking work logs, extract key phrases to improve picking efficiency, and suggest the optimal picking order in real time."

[1138] This system makes it possible to optimize the efficiency of sales activities as well as picking operations at logistics facilities in real time.

[1139] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1140] Step 1:

[1141] The server collects past transaction data and picking data. It extracts data from the transaction database and picking database as input and obtains data converted into text format as output. Specifically, it converts data such as minutes, transaction records, email correspondence, and picking logs into a format that is easy to centrally manage.

[1142] Step 2:

[1143] The server analyzes the collected transaction and picking data to extract key phrases related to success and failure. It uses the collected text data as input and generates a list of key phrases as output. Specifically, it vectorizes the text data using TfidfVectorizer and identifies important key phrases using RandomForestClassifier.

[1144] Step 3:

[1145] The server generates and updates models that predict the likelihood of agreement in negotiations and the efficiency of picking work based on the extracted key phrases. It uses the key phrase list as input and generates a predictive model as output. Specifically, it analyzes the frequency and relationships of key phrases and builds the model using regression analysis and machine learning algorithms.

[1146] Step 4:

[1147] The device identifies key phrases in real time during negotiations or product picking, and issues warnings or suggestions as needed. It uses real-time negotiation voice data or picking logs as input and generates warning or suggestion messages as output. Specifically, it works in conjunction with voice recognition software (e.g., Google Speech-to-Text) to detect pre-defined key phrases in real time.

[1148] Step 5:

[1149] After a sales negotiation or picking task is completed, the user provides feedback to the system. The results of the negotiation (success or failure) or the results of the picking task (achievement rate, number of errors) are input to the system, and feedback data is generated as output. Specifically, the results are entered through a mobile application or web interface and sent to the server as feedback data.

[1150] Step 6:

[1151] The server retrains the predictive model based on the feedback to improve its accuracy. It uses the feedback data as input and generates an improved predictive model as output. Specifically, it feeds the newly collected feedback data into a machine learning algorithm to optimize the model's parameters.

[1152] 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.

[1153] The present invention relates to a system that uses a generation AI and an emotion engine to systematize know-how for reaching agreements in sales negotiations, and specific embodiments thereof are described below.

[1154] 1. Data Collection

[1155] The server first collects past transaction data from the company's internal database, including meeting minutes, transaction records, and email correspondence, and then converts the data into text format for analysis.

[1156] Specific examples

[1157] The server extracts business meeting minutes from January 2022 to January 2023 from the database, converts them into text format, and saves them.

[1158] 2. Keyphrase Extraction

[1159] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques.

[1160] Specific examples

[1161] The server analyzes past sales negotiation data and identifies key phrases such as "ROI (return on investment)," "implementation cost," and "customer support."

[1162] 3. Creating and updating the model

[1163] The server generates and updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model includes calculation methods and algorithms for predicting the success or failure of a business negotiation.

[1164] Specific examples

[1165] The server creates a model that predicts the likelihood of agreement using regression analysis and machine learning models based on key phrases that frequently appear in sales negotiations with a success rate of 80% or more.

[1166] 4. Real-time support for business negotiations

[1167] The terminal (salesperson's device) identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the key phrase list provided by the server.

[1168] Specific examples

[1169] If the key phrase "customer support" appears during a sales negotiation, the terminal notifies the sales representative that the likelihood of an agreement increases.

[1170] When entering minutes of a business meeting, the terminal displays a warning if an important key phrase is omitted.

[1171] 5. Leveraging Emotional Engines

[1172] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[1173] Specific examples

[1174] The device uses a camera to analyze the customer's facial expressions during sales negotiations and notifies the sales representative if emotions such as anxiety or excitement are detected.

[1175] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction and makes suggestions to adjust the progress of the business negotiations.

[1176] 6. Feedback on sales results

[1177] After the sales negotiation, the user (salesperson) provides feedback to the system, and the server uses this feedback to further refine the prediction model and emotion recognition model.

[1178] Specific examples

[1179] The user inputs feedback into the system as a result of the business negotiation, such as "success" or "failure."

[1180] The server analyzes this feedback data and retrains the model to improve its predictive accuracy.

[1181] This system allows salespeople to grasp important key phrases and the emotional state of the customer during the first sales meeting, increasing the likelihood of an agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[1182] The processing flow will be explained below.

[1183] Step 1: Data collection

[1184] The server extracts past transaction data from the company's internal database, including business meeting minutes, transaction records, and email correspondence. The collected data is then converted into text format, making it suitable for analysis.

[1185] Specific actions

[1186] The server automatically extracts transaction data from the database every night.

[1187] The server converts the extracted data into text format and prepares it into an analyzable file format.

[1188] Step 2: Extracting Keyphrases

[1189] The server analyzes the collected transaction data with generative AI to extract key phrases related to success and failure. This analysis utilizes machine learning and natural language processing (NLP) techniques.

[1190] Specific actions

[1191] The server inputs the collected transaction data into a machine learning model and performs NLP analysis.

[1192] The server compares data from successful and unsuccessful transactions to identify frequently occurring key phrases.

[1193] Step 3: Generate and update the model

[1194] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically.

[1195] Specific actions

[1196] The server creates a predictive model using regression analysis and machine learning algorithms based on key phrases that frequently appear in business negotiations with a high success rate.

[1197] The server periodically retrains the model with newly collected data to improve prediction accuracy.

[1198] Step 4: Real-time support for business negotiations

[1199] The terminal, as a device for sales representatives, identifies key phrases in real time during sales negotiations, provides warnings and suggestions as necessary, and evaluates the progress of sales negotiations based on the latest key phrase list provided by the server.

[1200] Specific actions

[1201] The device uses voice recognition technology to analyze speech during negotiations and notify the sales representative when important key phrases appear.

[1202] The terminal displays a warning if an important key phrase is missing when entering business meeting minutes.

[1203] Step 5: Leverage the Emotion Engine

[1204] The device uses an emotion engine to analyze the user's (customer's) emotional state in real time. The emotion engine recognizes emotions by analyzing multiple data points such as the user's tone of voice, facial expressions, and behavior.

[1205] Specific actions

[1206] The device uses a camera to analyze the customer's facial expressions during the sales negotiation and recognizes their emotional state (e.g., anxiety, excitement, satisfaction, etc.) in real time.

[1207] The device uses voice tone analysis to detect the customer's level of interest or dissatisfaction, providing the sales representative with information to adjust the progress of the negotiation and the content of the proposal.

[1208] Step 6: Feedback on the outcome of the deal

[1209] After a sales meeting, the user (sales representative) provides feedback to the system, and the server uses this feedback to improve the accuracy of the prediction model and emotion engine.

[1210] Specific actions

[1211] The user enters the outcome of the negotiation (success or failure) and the reason for it into the system.

[1212] The server analyzes the feedback data and retrains the predictive model and emotion engine to improve accuracy.

[1213] Example 2

[1214] 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."

[1215] To increase the likelihood of agreement in sales negotiations, it is important to utilize past transaction data to identify factors related to success and failure. However, conventional methods make it difficult to efficiently analyze these factors and support sales negotiations in real time. In addition, more effective sales negotiation support is required by understanding the customer's emotional state, but there is a lack of technology to perform emotional analysis in real time. To solve these issues, a comprehensive and efficient system is required.

[1216] 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.

[1217] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data and extracting key phrases related to success and failure, means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases, means for identifying key phrases in real time during a sales negotiation and issuing a warning or suggestion, means for analyzing the emotional state of the customer in real time using a sentiment analysis engine and notifying the sales representative, and means for collecting the results of the sales negotiation as feedback and improving the accuracy of the model. This makes it possible to accurately grasp important factors in sales negotiations and increase the likelihood of agreement in sales negotiations.

[1218] "Transaction data" refers to a series of records related to business negotiations, such as minutes of past business negotiations, transaction records, and email exchanges.

[1219] "Key phrases" refer to important words or short phrases related to the success or failure of a business deal, extracted through analysis of transaction data.

[1220] An "agreement probability prediction model" refers to a mathematical model that includes calculation methods and algorithms for predicting the success or failure of a business negotiation based on extracted key phrases.

[1221] "Generative AI" refers to artificial intelligence that uses machine learning and natural language processing techniques to analyze collected transaction data and extract key phrases.

[1222] "Feedback" refers to the information on the results of sales negotiations that sales representatives enter into the system after the negotiations are completed, and is data used to improve the accuracy of the model.

[1223] An "emotion analysis engine" refers to technology that analyzes data such as a customer's tone of voice and facial expressions to recognize their emotional state in real time.

[1224] "Real-time analysis" refers to a processing method that processes data instantly while a sales negotiation is in progress and provides sales representatives with the necessary information and warnings immediately.

[1225] "Salesperson's device" refers to a device (e.g., PC, tablet, smartphone) that receives real-time information and displays alerts and suggestions during a sales meeting.

[1226] The present invention relates to a system for systematizing sales negotiation know-how using a generation AI and a sentiment analysis engine. Specific embodiments of the system are described below.

[1227] First, the server collects past transaction data from the company's internal database. This transaction data includes sales meeting minutes, transaction records, and email correspondence, and converts it into text format and saves it. For example, the server uses an SQL query to extract sales data from January 2022 to January 2023, and then uses a Python script to save it as a text JSON file.

[1228] Next, the server analyzes the collected transaction data using a generative AI (e.g., GPT-4) to extract key phrases related to success and failure. This analysis is performed using machine learning and natural language processing (NLP) techniques. Specifically, the server preprocesses the collected text data using a natural language processing library (e.g., spaCy, NLTK), inputs it into the generative AI to extract key phrases, and stores the extracted key phrases in a database.

[1229] The server then generates and updates a model to predict the likelihood of agreement based on the extracted key phrases. This model includes algorithms (e.g., regression analysis, machine learning) for predicting the success or failure of a deal. Specifically, the server combines the extracted key phrases with the deal outcome data, trains a regression model or classifier using a machine learning library (e.g., scikit-learn), and stores the trained model in a model repository on the server.

[1230] The terminal (sales representative's device) identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The progress of the negotiation is evaluated based on the key phrase list provided by the server. Specifically, the terminal uses a real-time data analysis module to convert the conversation during the negotiation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears. It also issues a warning if an important key phrase is omitted when creating minutes of the negotiation.

[1231] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The emotion analysis engine recognizes emotions by analyzing data points such as the customer's voice tone and facial expressions. Specifically, the device captures the customer's facial expressions with a camera, performs image analysis using a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[1232] Finally, after the sales negotiation, the user (salesperson) provides feedback to the system on the outcome. The server uses this feedback to further refine the prediction model and emotion recognition model. Specifically, the user enters "success" or "failure" as the outcome of the negotiation into the system, and the server uses this feedback data as a new learning dataset to retrain the model.

[1233] Example prompt

[1234] 1. "Extract key phrases related to success and failure from past transaction data."

[1235] 2. "Suggest a way to analyze a customer's tone of voice and facial expressions to signal their emotional state during a sales conversation."

[1236] 3. "How can we refine our predictive models using data from sales rep feedback on deal outcomes?"

[1237] By utilizing this system, sales representatives can accurately grasp the important elements of sales negotiations, increasing the likelihood of agreement. It also reduces unnecessary approaches and enables rational and efficient sales activities.

[1238] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1239] Step 1:

[1240] The server collects past transaction data from the company's internal database. Input data includes meeting minutes, transaction records, and email correspondence. This data is converted into text format and output. Specifically, the server uses SQL queries to extract the transaction data, converts it into a text-format JSON file using a Python script, and saves it.

[1241] Step 2:

[1242] The server analyzes the collected transaction data using a generation AI and extracts key phrases related to success and failure. The input is text-format transaction data, and the output is the extracted key phrases. Specifically, the server performs preprocessing using a natural language processing library (e.g., spaCy, NLTK), inputs the data into the generation AI to extract key phrases, and stores them in a database.

[1243] Step 3:

[1244] The server generates and updates a model that predicts the probability of agreement based on the extracted key phrases. The inputs are key phrases and negotiation outcome data, and the output is a predictive model. Specifically, the server creates a training dataset using a machine learning library (e.g., scikit-learn), trains a regression model or classifier, and saves the trained model in a model repository.

[1245] Step 4:

[1246] The terminal identifies key phrases in real time during negotiations and issues warnings or suggestions as necessary. The input is conversation data from the negotiation, and the output is the identified key phrases and warnings or suggestions based on them. Specifically, the terminal uses a real-time data analysis module to convert the conversation into text, compares it with the server's key phrase list, and displays an alert if a matching key phrase appears.

[1247] Step 5:

[1248] The device uses an emotion analysis engine to analyze the user's (customer's) emotional state in real time. The inputs are voice tone and facial expression data, and the output is the analyzed emotional state. Specifically, the device captures the customer's facial expression with a camera, analyzes it with a facial expression recognition library (e.g., OpenCV), records audio with a microphone, analyzes the emotion using a voice tone analysis library (e.g., Librosa), and notifies the sales representative of the results.

[1249] Step 6:

[1250] After a sales negotiation, the user (salesperson) feeds the results back to the system. The input is the negotiation result (success or failure), and the output is an updated prediction model. Specifically, the user inputs the negotiation result into the system, and the server uses this feedback data to retrain the model, and retrains it to improve the model's accuracy.

[1251] (Application example 2)

[1252] 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."

[1253] The progress of sales negotiations and consensus building in sales activities often depend on the experience and intuition of individual sales representatives, resulting in variations in the success rate of sales negotiations. Furthermore, it is not easy to appropriately grasp and respond to the customer's emotional state, and effective methods are needed to increase the likelihood of agreement in sales negotiations. Furthermore, systems that support sales negotiations in real time lack accuracy and practicality, so the development of more advanced and effective sales negotiation support systems is necessary.

[1254] 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.

[1255] In this invention, the server includes means for collecting past transaction data, means for analyzing the collected transaction data to extract key phrases related to success and failure, and means for generating and updating a model for predicting the likelihood of agreement based on the extracted key phrases. This allows sales representatives to identify the key phrases in real time during negotiations and receive warnings or suggestions. Furthermore, by including emotion analysis means for analyzing the emotional state of the customer and means for making suggestions for adjusting the progress of the negotiations based on the emotional data analyzed by the emotion analysis means, appropriate responses can be made based on the customer's emotions, further increasing the likelihood of agreement in the negotiations. Furthermore, by including means for collecting the results of the negotiations as feedback and improving the accuracy of the model, the accuracy of the system can be continuously improved.

[1256] "Transaction data" refers to all information related to past business negotiations and negotiations, including minutes of business negotiations, transaction records, and email correspondence.

[1257] "Collection means" is a general term for the equipment and programs used to compile the necessary data and gather information.

[1258] "Analysis tools" is a general term for the equipment and algorithms used to examine collected data in detail and extract meaningful information from it.

[1259] "Key phrases" refer to important phrases or terms related to the success or failure of a deal.

[1260] "Agreement likelihood" refers to the likelihood that a particular deal will be successful and result in an agreement.

[1261] "Emotion analysis means" is a general term for devices and programs that analyze a customer's tone of voice, facial expressions, movements, etc., to identify the customer's emotional state.

[1262] "Real-time identification means" is a general term for devices and programs that perform real-time data analysis during business negotiations and instantly identify important information.

[1263] "Feedback tools" is a general term for devices and programs that systematically collect results obtained after business negotiations and are used to improve and learn from the system.

[1264] This invention is a system that uses a generation AI and an emotion engine to systematize sales negotiation agreement know-how in sales activities, and specific embodiments will be described below.

[1265] 1. Data Collection and Preprocessing

[1266] The server first collects data on past business negotiations and transactions from the company's internal database, including minutes of business negotiations, transaction records, and email correspondence, and then converts the collected data into text format for analysis.

[1267] 2. Keyphrase Extraction

[1268] The server analyzes the collected transaction data using generative AI to extract key phrases related to success and failure. This analysis uses machine learning and natural language processing (NLP) techniques. Specific tools include Hugging Face's Transformers.

[1269] 3. Creating and updating the model

[1270] The server generates and periodically updates a model that predicts the likelihood of agreement based on the extracted key phrases. This model uses regression analysis or machine learning algorithms, such as open-source libraries like scikit-learn.

[1271] 4. Real-time support for business negotiations

[1272] The salesperson's device (e.g., smart glasses) identifies key phrases in real time during a sales call and provides warnings or suggestions as needed. The device immediately notifies the customer when key phrases such as "warranty period" or "additional services" appear during the call. Software such as OpenCV and SpeechRecognition are also used to analyze the customer's facial expressions and tone of voice and suggest appropriate responses based on their emotional state.

[1273] 5. Leveraging Emotional Engines

[1274] The device uses an emotion engine to analyze the customer's emotional state and provide feedback to the salesperson. For example, if the customer looks anxious, the device will use voice tone analysis to suggest appropriate countermeasures.

[1275] 6. Feedback on sales results

[1276] After a sales meeting is completed, the sales representative provides feedback to the system, and the server uses this feedback to further improve the accuracy of the prediction model and emotion recognition model.

[1277] Specific examples

[1278] For example, if a salesperson wearing smart glasses is asked "How long is the warranty period?" during a sales negotiation with a customer in a physical store, the system will detect the important key phrase "warranty period." If the customer looks anxious, the system will notify the salesperson to provide a more detailed explanation, thereby helping to increase the likelihood of an agreement.

[1279] Prompt Sentence Examples

[1280] When a customer asks, "How long is the product warranty period?", the system detects the important key phrase "warranty period" and notifies the sales representative. If anxiety is detected from the customer's facial expression, the system suggests a solution.

[1281] This allows salespeople to respond optimally to the needs and emotional state of the customer, thereby improving the success rate of sales negotiations.

[1282] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1283] Flow of the system program that realizes the application example

[1284] Step 1:

[1285] The server collects past transaction data from the company's internal database and converts it into text format. The collected data includes minutes of business meetings, transaction records, and email correspondence. It receives data from the company's internal database as input and outputs text data that can be analyzed.

[1286] Step 2:

[1287] The server analyzes the collected text data using generative AI and natural language processing (NLP) techniques to extract key phrases related to the success and failure of business negotiations. It receives text data as input and obtains key phrases related to success and failure as output. This processing uses Hugging Face's Transformers library.

[1288] Step 3:

[1289] The server generates a model that predicts the likelihood of agreement based on the extracted key phrases and updates it periodically. Specifically, it uses regression analysis and machine learning algorithms. It receives key phrases as input and obtains a predictive model as output. Machine learning libraries such as scikit-learn are used.

[1290] Step 4:

[1291] The device (e.g., smart glasses) captures audio and video in real time during a sales meeting and analyzes it. Specifically, the camera captures the customer's facial expressions and the microphone records their voice. The real-time data obtained from the camera and microphone is received as input, and the analysis results are obtained as output. OpenCV and the SpeechRecognition library are used.

[1292] Step 5:

[1293] The device uses the generative AI model from the server to identify key phrases that appear during sales negotiations in real time and provide warnings or suggestions to the sales representative as needed. It receives real-time analysis results and predictive models as input and provides notifications and suggestions as output.

[1294] Step 6:

[1295] The device uses an emotion analysis engine to analyze the voice tone and facial expression data during the negotiation to detect the customer's emotional state. It receives audio and video data as input and outputs the analysis result of the customer's emotional state.

[1296] Step 7:

[1297] The device then proposes appropriate countermeasures to the salesperson based on the results of the emotion analysis. For example, if the customer is feeling anxious, it will notify them to provide a detailed explanation. The device receives the emotion analysis results as input and proposes countermeasures as output.

[1298] Step 8:

[1299] After the negotiation is completed, the user provides feedback on the outcome (success or failure) to the system, and the server retrains the prediction model based on this data. The server receives the negotiation result data as input and obtains an improved prediction model as output.

[1300] 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.

[1301] 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.

[1302] 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.

[1303] 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.

[1304] FIG. 9 is a diagram illustrating 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 actions 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.

[1305] 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.

[1306] 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).

[1307] 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.

[1308] 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."

[1309] 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.

[1310] 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).

[1311] 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.

[1312] 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.

[1313] 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.

[1314] 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.

[1315] 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.

[1316] 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.

[1317] 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.

[1318] 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.

[1319] 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.

[1320] 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.

[1321] The following is further disclosed regarding the above embodiment.

[1322] (Claim 1)

[1323] a means of collecting historical transaction data;

[1324] means for analyzing the collected transaction data and extracting key phrases related to success and failure;

[1325] A means for generating and updating a model for predicting agreement probability based on the extracted key phrases;

[1326] means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions;

[1327] A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model;

[1328] A system including:

[1329] (Claim 2)

[1330] 10. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

[1331] (Claim 3)

[1332] 2. The system according to claim 1, wherein the generated key phrase list is sent to a sales negotiation application and displayed on a sales representative's terminal.

[1333] "Example 1"

[1334] (Claim 1)

[1335] a means of collecting historical transaction data;

[1336] means for analyzing the collected transaction data and extracting key phrases related to success and failure;

[1337] A means for generating and updating a model for predicting agreement probability based on the extracted key phrases;

[1338] means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions;

[1339] A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model;

[1340] means for converting said transaction data into a text format using optical character recognition technology;

[1341] a means for analyzing transaction data using a generative AI model and inputting prompt text;

[1342] A system including:

[1343] (Claim 2)

[1344] 10. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

[1345] (Claim 3)

[1346] 2. The system according to claim 1, wherein the generated key phrase list is sent to a sales negotiation application and displayed on a sales representative's terminal.

[1347] "Application Example 1"

[1348] (Claim 1)

[1349] a means of collecting historical transaction data;

[1350] means for analyzing the collected transaction data and extracting key phrases related to success and failure;

[1351] A means for generating and updating a model for predicting agreement probability based on the extracted key phrases;

[1352] means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions;

[1353] A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model;

[1354] A means for applying the model to support operational efficiency in a logistics facility and making real-time suggestions when a robot performs product picking operations;

[1355] A system including:

[1356] (Claim 2)

[1357] 10. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

[1358] (Claim 3)

[1359] The system of claim 1, wherein the generated key phrase list is sent to a sales negotiation application or a picking optimization application and displayed on a terminal of a sales representative or robot.

[1360] "Example 2: Combining Emotion Engines"

[1361] (Claim 1)

[1362] a means of collecting historical transaction data;

[1363] means for analyzing the collected transaction data and extracting key phrases related to success and failure;

[1364] A means for generating and updating a model for predicting agreement probability based on the extracted key phrases;

[1365] means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions;

[1366] A means of analyzing the emotional state of customers in real time using a sentiment analysis engine and notifying sales representatives;

[1367] A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model;

[1368] A system including:

[1369] (Claim 2)

[1370] 10. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

[1371] (Claim 3)

[1372] 2. The system according to claim 1, wherein the generated key phrase list is sent to a sales negotiation application and displayed on a sales representative's terminal.

[1373] "Application example 2 when combining emotion engines"

[1374] (Claim 1)

[1375] a means of collecting historical transaction data;

[1376] means for analyzing the collected transaction data and extracting key phrases related to success and failure;

[1377] A means for generating and updating a model for predicting agreement probability based on the extracted key phrases;

[1378] means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions;

[1379] emotion analysis means for analyzing the emotional state of a customer;

[1380] a means for making a proposal to adjust the progress of the business negotiation based on the emotion data analyzed by the emotion analysis means;

[1381] A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model;

[1382] A system including:

[1383] (Claim 2)

[1384] 10. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

[1385] (Claim 3)

[1386] 2. The system according to claim 1, wherein the generated key phrase list is sent to a sales negotiation application and displayed on a sales representative's terminal. [Explanation of symbols]

[1387] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means of collecting historical transaction data; means for analyzing the collected transaction data and extracting key phrases related to success and failure; A means for generating and updating a model for predicting agreement probability based on the extracted key phrases; means for identifying said key phrases in real time during a business negotiation and providing warnings or suggestions; A means for collecting the results of the business negotiations as feedback and improving the accuracy of the model; A system including:

2. The system of claim 1, wherein the key phrases are identified by comparing data on successful transactions with data on unsuccessful transactions.

3. 2. The system according to claim 1, wherein the generated key phrase list is sent to a sales negotiation application and displayed on a sales representative's terminal.

Images