Intelligent secretary autonomous learning and memory large model tool application

Abstract

The invention belongs to the technical field of AI large models, and particularly relates to an intelligent secretary autonomous learning and memory large model tool application which is used for understanding inquiry and commands of clients and achieving natural language inquiry, voice assistant functions and efficient client service interaction. According to historical behaviors, browsing records and purchase records of the user, a user portrait is generated, personalized commodity recommendation is provided, and the conversion rate is increased; the description, price and inventory information of commodities are managed and updated, and it is ensured that latest commodity data are displayed to users; the user is assisted in order placing and payment, and tracking, confirmation and exception handling tasks of the order are processed; the interaction history of the user is stored and analyzed to help formulate a marketing strategy and provide personalized customer support; users are helped to know commodity inventory conditions and logistics tracking information, and inventory management and distribution processes are optimized; and user feedback and product evaluation are collected and analyzed, so that the product quality and the user experience are improved, and the effect of intelligently supervising the e-commerce platform is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of AI big models, and specifically to an application of a tool for intelligent secretary autonomous learning and memory of big models. Background Art

[0002] In the field of intelligent secretaries, with the development of artificial intelligence technology, intelligent secretaries have gradually become an indispensable part of people's daily life and work. They can provide information query, schedule management, reminders and other services according to user needs. However, traditional intelligent secretaries rely on pre-written rules and fixed databases to work, which makes them lack flexibility and adaptability when facing new information and unforeseen events.

[0003] In recent years, with the development of large model technology, intelligent secretaries have been given the ability to learn autonomously by imitating the way humans learn. This technology can refine knowledge and learn language models by analyzing large amounts of text data without human intervention, thereby improving their work efficiency and service quality. However, there are still some challenges in the existing autonomous learning and memory models, such as the model training process is complex and time-consuming, the learned knowledge needs to be continuously optimized and updated, and how to efficiently manage and apply a large amount of learning results in practical applications still requires further research and exploration.

[0004] In the field of e-commerce platforms, with the development of intelligent technology, user needs are becoming more diverse and complex. In the existing technology, although there are some intelligent customer service systems, they mainly rely on preset rules and service templates and cannot effectively handle a large amount of unstructured information and sudden personalized needs. In addition, the existing intelligent secretary system has limitations in learning and memorizing user behavior and transaction data, making it difficult to achieve efficient and accurate knowledge accumulation and transfer.

[0005] In order to improve the user experience and service efficiency of e-commerce platforms, existing technologies urgently need a large model tool that can autonomously learn and memorize user behavior, transaction data, and merchant feedback information. This tool can dynamically generate personalized service plans, provide personalized recommendations, and effectively handle various needs raised by users, including but not limited to product recommendations, price consultation, logistics tracking, after-sales service, etc., thereby significantly improving customer satisfaction and sales conversion rate. Through deep learning and natural language processing technology, the tool can also achieve natural dialogue with users, simulate the service style of human customer service, provide a humanized interactive experience, and further optimize the service quality and user stickiness of e-commerce platforms.

[0006] In view of the above-mentioned technical defects, an application solution of an intelligent secretary autonomous learning and memory large model tool is proposed. Summary of the invention

[0007] To solve the above problems, the present invention provides the following technical solutions:

[0008] An intelligent secretary autonomous learning and memory large model tool application, comprising:

[0009] Natural language processing module, which is used to understand customer inquiries and commands, and helps realize natural language query, voice assistant function and efficient customer service interaction;

[0010] User portrait and personalized recommendation module, which generates user portraits and provides personalized product recommendations based on the user's historical behavior, browsing history, and purchase history to improve conversion rate;

[0011] Product information management module manages and updates product descriptions, prices, and inventory information to ensure that the latest product data is displayed to users;

[0012] The order and payment management module assists users in placing orders and making payments, and handles order tracking, confirmation, and exception handling tasks;

[0013] Customer relationship management module, which stores and analyzes user interaction history to help formulate marketing strategies and provide personalized customer support;

[0014] Inventory and logistics management module helps users understand product inventory and logistics tracking information, and optimize inventory management and delivery processes;

[0015] Feedback and evaluation analysis module collects and analyzes user feedback and product evaluations to improve product quality and enhance user experience.

[0016] Furthermore, the natural language processing module is used to identify the intention of user input, including asking for product information, placing orders, checking order status, and obtaining discount information, to implement the basis for context-aware interaction, extract key entities in user input, manage the dialogue state between the user and the system, ensure context consistency in multiple rounds of dialogue, provide a continuous user experience, convert the user's voice input into text, and convert the system's text replies into voice output to support the voice assistant function, implement automatic answers, extract information from structured and unstructured data sources, quickly answer users' common questions, analyze whether the user's input contains emotional tendencies, and help the system adjust the tone or strategy of the response to improve user satisfaction.

[0017] Furthermore, the user portrait and personalized recommendation module collects data from multiple channels such as the user's browsing history, purchase history, search behavior, clicks and dwell time, uses data mining technology to analyze user behavior patterns, extracts valuable information from them, generates a multi-dimensional portrait of the user based on the collected data, including demographic information, interest preferences, purchasing power, and shopping habits, uses machine learning algorithms to dynamically update the user portrait to reflect changes in user preferences, makes recommendations based on user similarity or product similarity, makes recommendations based on the characteristic attributes of the product matching the interests shown in the user portrait, and combines multiple algorithms to improve recommendation accuracy, such as combining collaborative filtering with content recommendation, adjusting recommended content in real time based on the user's current browsing and search behavior, personalizing the homepage display based on the user portrait and historical behavior, sending personalized product recommendations and discount information, providing recommendations for related or supplementary products, sorting search results based on user preferences, collecting user feedback on recommended content, optimizing the recommendation algorithm through clicks, purchases, and explicit ratings, and using A / B testing to evaluate the effectiveness of different recommendation strategies, ensuring that data collection and use comply with privacy laws and regulations, providing transparent data usage instructions, and providing users with preference and privacy setting options to allow users to actively manage their own personalized settings.

[0018] Furthermore, the product information management module provides an easy-to-use interface or API for entering basic information of products, including name, description, price, inventory, SKU, barcode, supports batch import and export of product information to quickly update large amounts of product data, allows products to be organized into different categories, and tags are added to products to improve the discoverability of products, supports a multi-level classification structure to facilitate user browsing and searching, tracks product inventory levels, ensures that inventory information is updated in real time, sets alarms and notification functions to remind administrators when inventory falls below a set threshold, and supports multiple pricing strategies, including discounts, promotional prices, member prices, and fixed prices. It can store price history for easy tracing and analysis, manage product images, videos and other media files to ensure that these resources are of high quality and load quickly, support batch uploading and optimization of media files, define product attributes and specifications, support multi-specification product management, automatically generate SKU information for each specification, provide product search engine optimization tools to help improve rankings in search engine results, support product keyword management and optimization, implement content review mechanism to ensure that product descriptions and media comply with platform standards and legal requirements, support manual and automatic review processes, save historical versions of product information, and allow rollback and comparison of different versions for easy maintenance and review.

[0019] Furthermore, the order and payment management module supports generating orders from shopping carts, and allows modification of order information under certain conditions, provides order confirmation and preview functions, ensures that users verify information before submission, manages the entire life cycle of orders, including pending payment, paid, processing, shipped, completed, cancelled and returned status, updates order status in real time and notifies users, provides interfaces for users and administrators to query order status and history records, supports multi-dimensional search and filtering of orders, supports users to apply for order cancellation or return and exchange online, provides corresponding review and processing procedures, automatically handles return logistics and refund procedures, ensures user experience, supports multiple payment methods, including credit cards, debit cards, e-wallets (, bank transfers and cash on delivery, and provides flexible payment Option combination and switching mechanism to ensure the security and accuracy of the payment process, including redirection to the payment platform, payment result callback, and success or failure notification, maintaining the user's payment credentials and receipts, implementing encryption technology to protect the user's payment information, complying with the payment card industry data security standards, monitoring and preventing payment fraud, ensuring transaction security, supporting the application of various promotional activities, coupons and discount codes, and automatically calculating discounts, providing the display and use of promotional information in the payment process, automatically generating financial statements, supporting daily reconciliation and monthly settlement with third-party payment platforms, realizing revenue and cost calculation for different business cycles, providing a simple refund process and multiple refund channels, and tracking refund status to ensure that user refund requests are quickly processed in compliance with the return and exchange policy.

[0020] Furthermore, the customer relationship management module integrates multiple channels to collect customer information, maintains basic customer information and their purchase history and interaction records, builds customer portraits based on customer behavior, preferences and demographic data, and implements customer segmentation in order to provide personalized services and marketing activities for different customer groups, provides customer support through multiple channels such as telephone, email, online chat and social media, integrates customer service systems to ensure unified management and recording of all customer interactions, implements chatbots and automatic reply systems to help quickly resolve common problems, provides a self-service portal for customers to find solutions on their own, collects and analyzes customer feedback, improves product and service quality, uses survey tools to evaluate customer satisfaction, and identifies Improve customer engagement, implement loyalty programs, encourage repeat purchases and customer retention, conduct regular customer care activities, enhance customer loyalty, monitor and analyze the effectiveness of various marketing activities and promotions, optimize marketing strategies, provide detailed reports and dashboards to help companies gain insight into customer trends, conduct in-depth analysis of customer data, identify trends, behavior patterns and potential business opportunities, use data visualization tools to generate reports, support decision making, track key performance indicators such as customer acquisition cost, conversion rate, customer lifetime value, provide real-time data analysis, and help identify business problems and improvement plans.

[0021] Furthermore, the inventory and logistics management module implements real-time monitoring of inventory to ensure that inventory quantities are updated immediately after sales, returns and replenishment, supports multi-warehouse management, tracks the inventory status of each warehouse, sets upper and lower inventory warning lines, automatically alarms and prompts replenishment when inventory falls below a certain level, uses an automatic replenishment system, optimizes inventory based on historical sales data and seasonal demand forecasts, regularly performs inventory counts to ensure that actual inventory is consistent with system records, provides inventory adjustment functions to facilitate the handling of lost, expired or returned items, manages inventory units, supports tracking and management of batch numbers and serial numbers, provides first-in-first-out and last-in-first-out inventory management strategies, optimizes shipping warehouse selection based on order details, automatically generates picking, packaging and shipping instructions, improves shipping efficiency, supports integration of multiple transportation methods and multiple logistics carriers, compares prices and service quality of different carriers, selects the best solution, provides real-time tracking of order logistics status, monitors every link in the transportation process at any time, supports automatic sending of logistics status update notifications to customers, manages return logistics processes, including analysis and processing of return reasons, optimizes reverse logistics processes, and reduces unnecessary cost losses.

[0022] Furthermore, the feedback and evaluation analysis module supports collecting customer feedback from channels, including website evaluations, mobile application reviews, emails, social media, customer service records, and automated feedback collection systems, such as sending emails or text messages requesting evaluations after an order is completed, designing and sending customized questionnaires to obtain detailed feedback from customers, supporting multiple question types, and automatically collecting and storing survey results, collecting and displaying customer evaluations of products, including ratings, text comments, and image / video comments, implementing an evaluation review mechanism, filtering inappropriate content, and ensuring the authenticity and objectivity of evaluations, collecting customer evaluations of merchants or services, helping other users understand the reputation of merchants, providing merchant reply functions, and promoting communication between merchants and customers, using natural language processing technology to perform sentiment analysis on customer feedback, identifying positive, negative, and neutral emotions, extracting common keywords and themes, identifying hot topics and potential problems for companies, analyzing the average ratings of products and services, monitoring rating trends, identifying reasons for rating fluctuations, and generating detailed reports to help companies understand changes in customer satisfaction and their driving factors, classifying and analyzing customer complaints, identifying high-frequency problems and areas that are likely to cause customer dissatisfaction, providing complaint handling process management, tracking handling progress, and ensuring timely resolution.

[0023] Furthermore, the data analysis includes collecting and establishing data from multiple dimensions of the user's data, extracting the data by cleaning the data, filtering the data in the database, filtering out the data in the database and establishing a new data set, including:

[0024]

[0025] Among them, Y(t) is the current comparison median data, (X(t)+X(t-1)+…+X(t-M+1)) is the sum of the data in the current database, and M is the number of data in the current database. By analyzing the output data at the current time point, calculating the median of all the data in the database according to the number, comparing the comparison median data with the current output data, and judging whether the current data meets the inclusion requirements of the data set according to the error range, the data is recorded.

[0026] Furthermore, the data visualization includes optimizing the analysis model by carrying out a loss function on the processed data. The loss function includes a contrast loss. The contrast loss optimizes the model so that the distances of similar sample pairs are close and the distances of dissimilar sample pairs are far away. For each pair of samples, the loss function calculates their distance in the embedding space and adjusts the model parameters according to the distance. The formula is as follows;

[0027] L=(1-y)·max(0,md)+y·d,

[0028] Among them, d is the distance between sample pairs, m is a set boundary value, y indicates whether the sample pair is a positive sample pair, 1 indicates a positive sample pair, and 0 indicates a negative sample pair. The contrast loss is calculated. For each query and its related data, its corresponding embedding vector is taken as a positive sample pair.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] In an application of an intelligent secretary self-learning and memorizing large model tool of the present invention, natural language query, voice assistant function and efficient customer service interaction are realized by understanding the inquiries and commands of customers; user portraits are generated and personalized product recommendations are provided according to the user's historical behavior, browsing history and purchase history, so as to improve the conversion rate; the description, price and inventory information of the products are managed and updated to ensure that the latest product data is displayed to the user; the user is assisted in placing orders and payments, and the tracking, confirmation and exception handling tasks of the orders are handled; the user's interaction history is stored and analyzed to help formulate marketing strategies and provide personalized customer support; the user is helped to understand the inventory status of the products and the logistics tracking information, and the inventory management and delivery processes are optimized; user feedback and product evaluations are collected and analyzed to improve product quality and enhance the user experience, and the effect of intelligent supervision of the e-commerce platform is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0032] Figure 1 It is an overall schematic diagram of an application of a large model tool for autonomous learning and memory of an intelligent secretary of the present invention;

[0033] Figure 2 A schematic diagram of a computer device for applying a large model tool of an intelligent secretary's autonomous learning and memory according to the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] like Figure 1-2 As shown, an intelligent secretary autonomous learning and memory large model tool application includes:

[0036] Natural language processing module, which is used to understand customer inquiries and commands, and helps realize natural language query, voice assistant function and efficient customer service interaction;

[0037] User portrait and personalized recommendation module, which generates user portraits and provides personalized product recommendations based on the user's historical behavior, browsing history, and purchase history to improve conversion rate;

[0038] Product information management module manages and updates product descriptions, prices, and inventory information to ensure that the latest product data is displayed to users;

[0039] The order and payment management module assists users in placing orders and making payments, and handles order tracking, confirmation, and exception handling tasks;

[0040] Customer relationship management module, which stores and analyzes user interaction history to help formulate marketing strategies and provide personalized customer support;

[0041] Inventory and logistics management module helps users understand product inventory and logistics tracking information, and optimize inventory management and delivery processes;

[0042] Feedback and evaluation analysis module collects and analyzes user feedback and product evaluations to improve product quality and enhance user experience.

[0043] Specifically, the natural language processing module is used to identify the intention of user input, including asking for product information, placing orders, checking order status, and obtaining discount information, to implement the basis for context-aware interaction, extract key entities in user input, manage the dialogue state between the user and the system, ensure context consistency in multiple rounds of dialogue, provide a continuous user experience, convert the user's voice input into text, and convert the system's text replies into voice output to support the voice assistant function, implement automatic answers, extract information from structured and unstructured data sources, quickly answer users' common questions, analyze whether the user's input contains emotional tendencies, and help the system adjust the tone or strategy of the response to improve user satisfaction.

[0044] Specifically, the user portrait and personalized recommendation module collects data from multiple channels such as the user's browsing history, purchase history, search behavior, clicks and dwell time, uses data mining technology to analyze user behavior patterns, extracts valuable information from them, generates a multi-dimensional portrait of the user based on the collected data, including demographic information, interest preferences, purchasing power, and shopping habits, uses machine learning algorithms to dynamically update the user portrait to reflect changes in user preferences, makes recommendations based on user similarity or product similarity, makes recommendations based on the characteristic attributes of the product matching the interests shown in the user portrait, and combines multiple algorithms to improve recommendation accuracy, such as combining collaborative filtering with content recommendation, adjusting recommended content in real time based on the user's current browsing and search behavior, personalizing the homepage display based on the user portrait and historical behavior, sending personalized product recommendations and discount information, providing recommendations for related or supplementary products, sorting search results based on user preferences, collecting user feedback on recommended content, optimizing the recommendation algorithm through clicks, purchases, and explicit ratings, and using A / B testing to evaluate the effectiveness of different recommendation strategies, ensuring that data collection and use comply with privacy laws and regulations, providing transparent data usage instructions, and providing users with preference and privacy setting options to allow users to actively manage their own personalized settings.

[0045] Specifically, the product information management module provides an easy-to-use interface or API for entering basic information about products, including name, description, price, inventory, SKU, and barcode, supports batch import and export of product information to quickly update large amounts of product data, allows products to be organized into different categories, and tags to be added to products to improve the discoverability of products, supports a multi-level classification structure to facilitate user browsing and searching, tracks product inventory levels, ensures that inventory information is updated in real time, sets alarms and notification functions to remind administrators when inventory falls below a set threshold, and supports multiple pricing strategies, including discounts, promotional prices, and member prices. Historical records are used for easy tracing and analysis. Product images, videos and other media files are managed to ensure that these resources are of high quality and load quickly. Bulk uploading and optimization of media files are supported. Product attributes and specifications are defined, and multi-specification product management is supported. SKU information for each specification is automatically generated. Product search engine optimization tools are provided to help improve rankings in search engine results. Product keyword management and optimization are supported. A content review mechanism is implemented to ensure that product descriptions and media comply with platform standards and legal requirements. Manual and automatic review processes are supported. Historical versions of product information are saved, and rollback and comparison of different versions are allowed for easy maintenance and review.

[0046] Specifically, the order and payment management module supports generating orders from shopping carts, and allows modifying order information under certain conditions, provides order confirmation and preview functions, ensures that users verify information before submission, manages the entire life cycle of orders, including pending payment, paid, processing, shipped, completed, cancelled and returned status, updates order status in real time and notifies users, provides interfaces for users and administrators to query order status and history records, supports multi-dimensional search and filtering of orders, supports users to apply for order cancellation or return and exchange online, provides corresponding review and processing procedures, automatically handles return logistics and refund procedures, ensures user experience, supports multiple payment methods, including credit cards, debit cards, e-wallets (, bank transfers and cash on delivery, provides flexible payment option combinations and switching mechanisms, ensures that the payment process is safe and correct, including redirection to payment The order and payment management module can be used to manage payment results, payment results callback, and success or failure notifications, maintain user payment credentials and receipts, implement encryption technology to protect user payment information, comply with payment card industry data security standards, monitor and prevent payment fraud, ensure transaction security, support the application of various promotional activities, coupons and discount codes, and automatically calculate discounts, provide display and use of promotional information in the payment process, automatically generate financial statements, support daily reconciliation and monthly settlement with third-party payment platforms, realize revenue and cost calculation for different business cycles, provide a simple refund process and multiple refund channels, and track refund status to ensure that user refund requests are quickly processed in compliance with the return and exchange policy. Through these functions, the order and payment management module promotes the improvement of user experience, ensures the security and smoothness of transactions, and simplifies the order and payment processing work in the background.

[0047] Specifically, the customer relationship management module integrates multiple channels to collect customer information, maintains basic customer information and their purchase history and interaction records, builds customer portraits based on customer behavior, preferences and demographic data, and implements customer segmentation in order to provide personalized services and marketing activities for different customer groups, provides customer support through multiple channels such as telephone, email, online chat and social media, integrates customer service systems to ensure unified management and recording of all customer interactions, implements chatbots and automatic reply systems to help quickly resolve common problems, provides a self-service portal for customers to find solutions on their own, collects and analyzes customer feedback, improves product and service quality, uses survey tools to evaluate customer satisfaction, and identifies Improve customer engagement, implement loyalty programs, encourage repeat purchases and customer retention, conduct regular customer care activities, enhance customer loyalty, monitor and analyze the effectiveness of various marketing activities and promotions, optimize marketing strategies, provide detailed reports and dashboards to help companies gain insight into customer trends, conduct in-depth analysis of customer data, identify trends, behavior patterns and potential business opportunities, use data visualization tools to generate reports, support decision making, track key performance indicators such as customer acquisition cost, conversion rate, customer lifetime value, provide real-time data analysis, and help identify business problems and improvement plans.

[0048] Implement data encryption measures to protect the security of customer information.

[0049] Ensure that the authentication infrastructure complies with relevant data privacy laws and regulations (such as GDPR).

[0050] Set up permission management to ensure that only authorized personnel can access and process customer data. Through these functions, the CRM module can help companies build stronger customer relationships, improve customer satisfaction and loyalty, and thus drive business growth.

[0051] Specifically, the inventory and logistics management module implements real-time monitoring of inventory to ensure that the inventory quantity is updated immediately after sales, returns and replenishment, supports multi-warehouse management, tracks the inventory status of each warehouse, sets upper and lower inventory warning lines, automatically alarms and prompts replenishment when the inventory is below a certain level, uses an automatic replenishment system, optimizes inventory based on historical sales data and seasonal demand forecasts, regularly performs inventory counts to ensure that the actual inventory is consistent with the system records, provides inventory adjustment functions to facilitate the handling of loss, expired or returned items, manages inventory units, supports tracking and management of batch numbers and serial numbers, provides first-in-first-out and last-in-first-out inventory management strategies, optimizes shipping warehouse selection based on order details, automatically generates picking, packaging and shipping instructions, improves shipping efficiency, supports the integration of multiple transportation methods and multiple logistics carriers, compares the prices and service quality of different carriers, selects the best solution, provides real-time tracking of order logistics status, monitors every link in the transportation process at any time, supports automatic sending of logistics status update notifications to customers, manages return logistics processes, including analysis and processing of return reasons, optimizes reverse logistics processes, and reduces unnecessary cost losses. Analyze inventory turnover, sales velocity, and slow-moving goods, and provide relevant reports.

[0052] Helps decide on inventory optimization measures to reduce excess inventory and out-of-stock situations.

[0053] Evaluate logistics operation indicators (such as on-time delivery rate, transportation costs and customer satisfaction).

[0054] Provides visual reports of key logistics data to support optimization and improvement of logistics networks.

[0055] System connection and integration: Integrate with supply chain management system, order management system and ERP system to achieve seamless data flow. Provide API interface to support integration with third-party logistics information system.

[0056] Security and privacy protection: Ensure the security of inventory and logistics data to prevent unauthorized access and data leakage. Implement encryption and access control measures to protect sensitive business information.

[0057] Through these functions, the inventory and logistics management module helps enterprises improve supply chain efficiency, reduce operating costs, and increase customer satisfaction, thereby gaining an advantage in market competition.

[0058] Specifically, the feedback and evaluation analysis module supports collecting customer feedback from channels, including website evaluations, mobile application reviews, emails, social media, customer service records, automated feedback collection systems, such as sending emails or text messages requesting evaluations after an order is completed, designing and sending customized questionnaires to obtain detailed customer feedback, supporting multiple question types, and automatically collecting and storing survey results, collecting and displaying customer evaluations of products, including ratings, text reviews, and image / video reviews, implementing evaluation review mechanisms, filtering inappropriate content, ensuring the authenticity and objectivity of evaluations, collecting customer evaluations of merchants or services, helping other users understand the reputation of merchants, providing merchant reply functions, and promoting communication between merchants and customers, using natural language processing technology to perform sentiment analysis on customer feedback, identifying positive, negative, and neutral sentiments, extracting common keywords and topics, identifying hot topics and potential problems for enterprises, analyzing the average ratings of products and services, monitoring rating trends, identifying reasons for rating fluctuations, generating detailed reports, helping enterprises understand changes in customer satisfaction and their driving factors, classifying and analyzing customer complaints, identifying high-frequency problems and areas that are prone to customer dissatisfaction, providing complaint handling process management, tracking handling progress, and ensuring timely resolution. Promote the optimization of product functions and features based on customer feedback and suggestions.

[0059] Use feedback analysis results to guide new product development and existing product improvements. Identify pain points in the customer journey and optimize the user experience. Implement specific improvement actions and verify the improvement effects through subsequent feedback.

[0060] Dashboard and report generation: Provides visual dashboards to display key feedback analysis indicators and customer satisfaction reports. Supports custom report generation for reference by senior management.

[0061] Feedback history and trend chart: Displays historical data and trend charts of feedback collection for long-term monitoring and comparison. Through these functions, the feedback and evaluation analysis module can help companies fully understand customer opinions, identify problems in a timely manner, and optimize products and services, thereby improving customer satisfaction and loyalty.

[0062] Specifically, the data analysis includes collecting and establishing data from multiple dimensions of the user, extracting data by cleaning the data, filtering the data in the database, filtering out the data in the database and establishing a new data set, including:

[0063]

[0064] Among them, Y(t) is the current comparison median data, (X(t)+X(t-1)+…+X(t-M+1)) is the sum of the data in the current database, and M is the number of data in the current database. By analyzing the output data at the current time point, calculating the median of all the data in the database according to the number, comparing the comparison median data with the current output data, and judging whether the current data meets the inclusion requirements of the data set according to the error range, the data is recorded.

[0065] Specifically, the data visualization includes optimizing the analysis model by carrying out a loss function on the processed data. The loss function includes a contrast loss. The contrast loss optimizes the model so that the distances of similar sample pairs are close and the distances of dissimilar sample pairs are far away. For each pair of samples, the loss function calculates their distance in the embedding space and adjusts the model parameters according to the distance. The formula is as follows;

[0066] L=(1-y)·max(0,md)+y·d,

[0067] Among them, d is the distance between sample pairs, m is a set boundary value, y indicates whether the sample pair is a positive sample pair, 1 indicates a positive sample pair, and 0 indicates a negative sample pair. The contrast loss is calculated. For each query and its related data, its corresponding embedding vector is taken as a positive sample pair.

[0068] According to another aspect of the present invention, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of applying the above-mentioned intelligent secretary autonomous learning and memory large model tool when executing the computer program. The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.;

[0069] Those skilled in the art will understand that Figure 2The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the computer device to which the present application scheme is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0070] The working principle of the application of the intelligent secretary self-learning and memory large model tool of the present invention is as follows: by understanding the inquiries and commands of customers, natural language query, voice assistant function and efficient customer service interaction are realized; according to the historical behavior, browsing history and purchase history of users, user portraits are generated and personalized commodity recommendations are provided to improve conversion rate; the description, price and inventory information of commodities are managed and updated to ensure that the latest commodity data is displayed to users; users are assisted in placing orders and payments, and order tracking, confirmation and exception handling tasks are handled; the user's interaction history is stored and analyzed to help formulate marketing strategies and provide personalized customer support; users are helped to understand the inventory status of commodities and logistics tracking information, and the inventory management and distribution processes are optimized; user feedback and product evaluations are collected and analyzed to improve product quality and enhance user experience, which has the effect of intelligently supervising e-commerce platforms.

[0071] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent secretary autonomous learning and memory large model tool application, characterized in that: include: Natural language processing module, which is used to understand customer inquiries and commands, and helps realize natural language query, voice assistant function and efficient customer service interaction; User portrait and personalized recommendation module, which generates user portraits and provides personalized product recommendations based on the user's historical behavior, browsing history, and purchase history to improve conversion rate; Product information management module manages and updates product descriptions, prices, and inventory information to ensure that the latest product data is displayed to users; The order and payment management module assists users in placing orders and making payments, and handles order tracking, confirmation, and exception handling tasks; Customer relationship management module, which stores and analyzes user interaction history to help formulate marketing strategies and provide personalized customer support; Inventory and logistics management module helps users understand product inventory and logistics tracking information, and optimize inventory management and delivery processes; Feedback and evaluation analysis module collects and analyzes user feedback and product evaluations to improve product quality and enhance user experience.

2. According to claim 1, the intelligent secretary autonomous learning and memory large model tool application is characterized in that: The natural language processing module is used to identify the intention of user input, including asking for product information, placing orders, checking order status, and obtaining discount information, to implement the basis of context-aware interaction, extract key entities in user input, manage the dialogue state between the user and the system, ensure context consistency in multiple rounds of dialogue, provide a continuous user experience, convert the user's voice input into text, and convert the system's text response into voice output to support the voice assistant function, implement automatic answers, extract information from structured and unstructured data sources, quickly answer users' common questions, analyze whether the user's input contains emotional tendencies, and help the system adjust the tone or strategy of the response to improve user satisfaction.

3. According to claim 2, the intelligent secretary autonomous learning and memory large model tool application is characterized in that: The user portrait and personalized recommendation module collects data from multiple channels such as the user's browsing history, purchase history, search behavior, clicks and dwell time, uses data mining technology to analyze user behavior patterns, extracts valuable information from them, generates a multi-dimensional portrait of the user based on the collected data, including demographic information, interest preferences, purchasing power, and shopping habits, uses machine learning algorithms to dynamically update the user portrait to reflect changes in user preferences, makes recommendations based on user similarity or product similarity, makes recommendations based on the characteristic attributes of the product matching the interests shown in the user portrait, and combines multiple algorithms to improve recommendation accuracy, such as combining collaborative filtering with content recommendation, adjusting recommended content in real time based on the user's current browsing and search behavior, personalizing the homepage display based on the user portrait and historical behavior, sending personalized product recommendations and discount information, providing recommendations for related or supplementary products, sorting search results based on user preferences, collecting user feedback on recommended content, optimizing the recommendation algorithm through clicks, purchases, and explicit ratings, and using A / B testing to evaluate the effectiveness of different recommendation strategies, ensuring that data collection and use comply with privacy laws and regulations, providing transparent data usage instructions, and providing users with preference and privacy setting options to allow users to actively manage their own personalized settings.

4. According to claim 3, the intelligent secretary autonomous learning and memory large model tool application is characterized in that: The product information management module provides an easy-to-use interface or API for entering basic information of products, including name, description, price, inventory, SKU, barcode, supports batch import and export of product information to quickly update large amounts of product data, allows products to be organized into different categories, and tags are added to products to improve the discoverability of products, supports a multi-level classification structure to facilitate user browsing and searching, tracks product inventory levels, ensures that inventory information is updated in real time, sets alarm and notification functions, and reminds administrators when inventory falls below a set threshold, supports multiple pricing strategies, including discounts, promotional prices, and member prices, and pricing history records for easy backtracking and analysis, manages product images, videos, and other media files, ensures that these resources are of high quality and load quickly, supports batch uploading and optimization of media files, defines product attributes and specifications, and supports multi-specification product management, automatically generates SKU information for each specification, provides product search engine optimization tools to help improve rankings in search engine results, supports product keyword management and optimization, implements a content review mechanism to ensure that product descriptions and media comply with platform standards and legal requirements, supports manual and automatic review processes, saves historical versions of product information, and allows rollback and comparison of different versions for easy maintenance and review.

5. According to claim 4, the intelligent secretary autonomous learning and memory large model tool application is characterized in that: The order and payment management module supports generating orders from shopping carts, and allows modifying order information under certain conditions, provides order confirmation and preview functions to ensure that users verify information before submission, manages the entire life cycle of orders, including pending payment, paid, processing, shipped, completed, cancelled and returned status, updates order status in real time and notifies users, provides interfaces for users and administrators to query order status and history records, supports multi-dimensional search and filtering of orders, supports users to apply for order cancellation or return and exchange online, provides corresponding review and processing procedures, automatically handles return logistics and refund procedures to ensure user experience, supports multiple payment methods, including credit cards, debit cards, e-wallets (, bank transfers and cash on delivery, and provides flexible payment options Combination and switching mechanisms to ensure the security and accuracy of the payment process, including redirection to the payment platform, payment result callback, and success or failure notification, maintaining users' payment credentials and receipts, implementing encryption technology to protect users' payment information, complying with payment card industry data security standards, monitoring and preventing payment fraud, ensuring transaction security, supporting the application of various promotional activities, coupons and discount codes, and automatically calculating discounts, providing display and use of promotional information in the payment process, automatically generating financial statements, supporting daily reconciliation and monthly settlement with third-party payment platforms, realizing revenue and cost calculation for different business cycles, providing a simple refund process and multiple refund channels, and tracking refund status to ensure that user refund requests are quickly processed in compliance with the return and exchange policy.

6. According to claim 5, the intelligent secretary autonomous learning and memory large model tool application is characterized in that: The customer relationship management module integrates multiple channels to collect customer information, maintains basic customer information and their purchase history and interaction records, builds customer portraits based on customer behavior, preferences and demographic data, and implements customer segmentation in order to provide personalized services and marketing activities for different customer groups, provides customer support through multiple channels such as telephone, email, online chat and social media, integrates customer service systems to ensure unified management and recording of all customer interactions, implements chatbots and automatic reply systems to help quickly resolve common problems, provides a self-service portal for customers to find solutions on their own, collects and analyzes customer feedback, improves product and service quality, uses survey tools to assess customer satisfaction, and identifies improvements. Improve customer engagement, implement loyalty programs, encourage repeat purchases and customer retention, conduct regular customer care activities, enhance customer loyalty, monitor and analyze the effectiveness of various marketing activities and promotions, optimize marketing strategies, provide detailed reports and dashboards, help companies gain insight into customer trends, conduct in-depth analysis of customer data, identify trends, behavior patterns and potential business opportunities, use data visualization tools to generate reports, support decision making, track key performance indicators such as customer acquisition cost, conversion rate, customer lifetime value, provide real-time data analysis, and help identify business problems and improvement plans.

7. The intelligent secretary autonomous learning and memory large model tool application according to claim 6 is characterized in that: The inventory and logistics management module implements real-time monitoring of inventory to ensure that inventory quantities are updated immediately after sales, returns and replenishment, supports multi-warehouse management, tracks the inventory status of each warehouse, sets upper and lower inventory warning lines, automatically alarms and prompts replenishment when inventory falls below a certain level, uses an automatic replenishment system, optimizes inventory based on historical sales data and seasonal demand forecasts, regularly performs inventory counts to ensure that actual inventory is consistent with system records, provides inventory adjustment functions to facilitate the handling of loss, expired or returned items, manages inventory units, supports tracking and management of batch numbers and serial numbers, provides first-in-first-out and last-in-first-out inventory management strategies, optimizes shipping warehouse selection based on order details, automatically generates picking, packaging and shipping instructions, improves shipping efficiency, supports integration of multiple transportation methods and multiple logistics carriers, compares prices and service quality of different carriers, selects the best solution, provides real-time tracking of order logistics status, monitors every link in the transportation process at any time, supports automatic sending of logistics status update notifications to customers, manages return logistics processes, including analysis and processing of return reasons, optimizes reverse logistics processes, and reduces unnecessary cost losses.

8. The intelligent secretary autonomous learning and memory large model tool application according to claim 7 is characterized in that: The feedback and evaluation analysis module supports collecting customer feedback from channels, including website evaluations, mobile application reviews, emails, social media, customer service records, and automated feedback collection systems, such as sending emails or text messages requesting evaluations after an order is completed, designing and sending customized questionnaires to obtain detailed feedback from customers, supporting multiple question types, and automatically collecting and storing survey results, collecting and displaying customer evaluations of products, including ratings, text reviews, and image / video reviews, implementing an evaluation review mechanism, filtering inappropriate content, and ensuring the authenticity and objectivity of evaluations, collecting customer evaluations of merchants or services, helping other users understand the reputation of merchants, providing merchant reply functions, and promoting communication between merchants and customers, using natural language processing technology to perform sentiment analysis on customer feedback, identifying positive, negative, and neutral sentiments, extracting common keywords and themes, identifying hot topics and potential problems for companies, analyzing the average ratings of products and services, monitoring rating trends, identifying reasons for rating fluctuations, and generating detailed reports to help companies understand changes in customer satisfaction and their driving factors, classifying and analyzing customer complaints, identifying high-frequency issues and areas that are likely to cause customer dissatisfaction, providing complaint handling process management, tracking handling progress, and ensuring timely resolution.

9. The intelligent secretary autonomous learning and memory large model tool application according to claim 8 is characterized in that: The data analysis includes collecting and establishing data from multiple dimensions of the user, extracting data by cleaning the data, filtering the data in the database, filtering out the data in the database and establishing a new data set, including: Among them, Y(t) is the current comparison median data, (X(t)+X(t-1)+…+X(t-M+1)) is the sum of the data in the current database, and M is the number of data in the current database. By analyzing the output data at the current time point, calculating the median of all the data in the database according to the number, comparing the comparison median data with the current output data, and judging whether the current data meets the inclusion requirements of the data set according to the error range, the data is recorded.

10. The intelligent secretary autonomous learning and memory large model tool application according to claim 8 is characterized in that: The data visualization includes optimizing the analysis model by carrying out a loss function on the processed data. The loss function includes a contrast loss. The contrast loss optimizes the model so that the distances of similar sample pairs are close and the distances of dissimilar sample pairs are far away. For each pair of samples, the loss function calculates their distance in the embedding space and adjusts the model parameters according to the distance. The formula is as follows; L=(1-y)·max(0,md)+y·d, Among them, d is the distance between sample pairs, m is a set boundary value, y indicates whether the sample pair is a positive sample pair, 1 indicates a positive sample pair, and 0 indicates a negative sample pair. The contrast loss is calculated. For each query and its related data, its corresponding embedding vector is taken as a positive sample pair.

Images