Cross-channel customer interaction management system and method based on data fusion

Abstract

The invention relates to the technical field of customer management, in particular to a cross-channel customer interaction management system and method based on data fusion. The method comprises the following steps: acquiring a customer communication demand, analyzing the customer communication demand, and determining customer basic information and customer communication content; based on multi-channel data, according to the basic information of the customer and the communication content of the customer, determining a consultation intention of the customer; and determining a reply thought according to the customer consultation intention, and responding to the customer according to a preset reply logic and the reply thought. User requirements are quickly responded, and the waiting time of customers is shortened. The personalized service is based on multi-channel data and communication history of the customer, and can provide services better meeting customer requirements, thereby improving customer satisfaction. And by utilizing machine learning and natural language processing technologies, the system can more accurately identify the consultation intention of the customer, and misunderstanding and repeated consultation are reduced. The accuracy of consultation response is improved, and the customer is ensured to obtain correct information and suggestions.

Description

Technical Field

[0001] The present application relates to the technical field of customer management, and in particular to a cross-channel customer interaction management system and method based on data fusion. Background Art

[0002] With the rapid development of information technology, the ways of interaction between enterprises and customers are becoming more and more diverse, such as communication through multiple channels such as telephone, email, social media, online chat, etc. This diversified interaction channel brings rich data resources, but also brings challenges to customer interaction management.

[0003] Traditional customer service and interaction management usually rely on a single communication channel and isolated customer information, making it difficult for companies to fully understand customer needs and behaviors. Therefore, how to effectively use these multi-channel data to improve customer service quality and increase customer satisfaction has become the focus of companies. Summary of the invention

[0004] The present application provides a cross-channel customer interaction management system and method based on data fusion to solve the above problems.

[0005] In a first aspect, the present application provides a cross-channel customer interaction management method based on data fusion, the method comprising: Obtain customer communication needs, analyze said customer communication needs, and determine customer basic information and customer communication content; Based on multi-channel data, determine the customer's consultation intention according to the customer's basic information and the customer's communication content; Determine the response ideas based on the customer's inquiry intention, and respond to the customer based on the preset response logic and the response ideas.

[0006] Through this solution, the burden of manual customer service is reduced and the efficiency of handling customer inquiries is improved through automated processing and intelligent response. Real-time data analysis enables the system to respond quickly to user needs and reduce customer waiting time. Personalized services are based on customers' multi-channel data and communication history, which can provide services that are more in line with customer needs and improve customer satisfaction. A consistent service experience means that customers can receive consistent service standards regardless of which channel they communicate through. Integrate multi-channel data to provide customers with comprehensive analysis reports to help companies better understand customer behavior. Using machine learning and natural language processing technology, the system can more accurately identify customers' consultation intentions and reduce misunderstandings and repeated consultations. Improve the accuracy of consultation responses to ensure that customers receive the correct information and advice.

[0007] Optionally, determining the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, includes: Analyze the customer communication content to determine whether the customer communication content contains product information; If included, determining whether the customer has channel information based on the multi-channel data and the customer basic information; If there is channel information, determining the order information of the customer according to the channel information; Matching the product information with the order information; Based on the matching results, the customer's consultation intention is determined.

[0008] Through this solution, the burden of manual customer service is reduced and the efficiency of handling customer inquiries is improved through automated processing and intelligent response. Customers are automatically classified according to their behavior and communication content to facilitate targeted services. Real-time data analysis enables the system to respond quickly to user needs and reduce customer waiting time. Personalized services are based on customers' multi-channel data and communication history, and can provide services that are more in line with customer needs and improve customer satisfaction. Through historical data, predict possible customer needs and prepare in advance. Provide personalized product or service recommendations based on customer preferences and purchase history. Consistent service experience, no matter which channel customers communicate through, they can get consistent service standards. Integrate multi-channel data to provide customers with comprehensive analysis reports to help companies better understand customer behavior and market trends.

[0009] Optionally, determining the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, includes: If the channel information does not include order information, determining the associated user based on the product information and the multi-channel data; Analyze the order records of the associated users to determine the reasons for the transactions; Determine the customer's consultation intention based on the reasons for the transaction and the content of the customer communication.

[0010] Through this solution, customized product recommendations and services are provided by analyzing customer historical data and communication content. Real-time analysis of customer communication provides immediate answers and help. Predict problems that customers may encounter and prepare in advance. Improve sales conversion rate through accurate consultation intention identification. Form a unified customer view by integrating data from different channels. Use machine learning algorithms to predict customer behavior and identify consultation intentions. Automated customer service is achieved through chatbots, automatic replies and other functions.

[0011] Optionally, determining the customer's consultation intention based on the transaction reason and the customer communication content includes: According to the transaction reason, retrieve the communication response record of the associated user; Based on the communication response records, retrieve the transaction time and operation records before the transaction; Determining whether the communication response record has a positive impact on the transaction based on the operation record and the transaction time; If so, the customer's consultation intention is determined based on the communication response record and the customer's consultation intention.

[0012] This solution can quickly identify and respond to customer inquiries, reducing waiting time. Provide personalized product or service recommendations based on customer historical behavior and inquiry content. Improve problem-solving efficiency by accurately identifying customer inquiry intentions. High-quality service experience helps improve customer retention. Use customer inquiry intention data to optimize marketing activities and increase conversion rates. Reduce the operating costs of the customer service department through automated processing.

[0013] Optionally, determining the associated user according to the product information and the multi-channel data includes: According to the product information, the multi-channel data is retrieved to determine whether there is a user who inquires about the product corresponding to the product information or purchases the product corresponding to the product information; If so, retrieve relevant information of the user who inquired about the product corresponding to the product information or purchased the product corresponding to the product information; Retrieving the purchase records of the multi-channel data according to the product information; Analyze the purchase records to determine important attributes of the purchaser; The relevant information is filtered according to the important attributes, and the associated users are determined according to the filtering results.

[0014] Through this solution, sales conversion rates can be improved through precision marketing for related users. By analyzing multi-channel data, customers who are interested in or have already purchased a specific product can be more accurately targeted. Quick response and personalized services for potential customers can shorten the sales cycle. Through multi-channel data, customer interaction patterns can be better understood and customer interaction can be enhanced. Based on purchase record analysis, inventory management can be optimized to reduce inventory backlogs. By analyzing customer needs, existing products can be improved or new products can be developed. For identified related users, precision marketing can be more effectively carried out to improve the conversion rate of marketing activities.

[0015] Optionally, determining the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, includes: If channel information exists, determining historical communication information of the customer based on the multi-channel data; Comparing the historical communication information with the customer communication content to determine whether the historical communication information contains related content that is consistent with the customer communication content; If so, retrieve and analyze the response record of the related content, and determine the direction of the customer's doubts based on the analysis results of the response record; According to the direction of the doubt, the customer's consultation intention is determined.

[0016] Through this solution, by identifying related users, enterprises can avoid repeatedly answering the same questions and reduce customer service workload. Analyzing customer communication history can help enterprises optimize service processes and reduce the occurrence of common problems. Personalized services and quick responses can improve customer loyalty, reduce customer churn, and provide high-quality customer service.

[0017] Optionally, determining whether the customer has channel information according to the customer basic information includes: Determine the identity information and login information of the customer based on the basic customer information; Matching the identity information and the login information with the multi-channel data respectively, and determining whether there is completely consistent channel data according to the matching results; If yes, it is determined that the customer has channel information; If not, analyzing the customer's communication content, and determining the customer's communication habits based on the analysis results; Analyzing whether there is channel data in the multi-channel data whose similarity to the communication habit reaches a preset similarity; If yes, the channel data, the identity information and the login information are analyzed to determine whether the customer has channel information.

[0018] Through this solution, through the matching and analysis of multi-channel data, enterprises can more accurately identify and associate multiple accounts and interactions of the same customer, thereby improving the integrity of customer identification. By comprehensively analyzing multi-channel data, enterprises can gain a deeper understanding of customers' purchasing behaviors, preferences, and needs, providing a basis for personalized marketing and services. Determining whether channel information exists for customers helps provide a consistent service experience, even if customers interact on different channels. By identifying customers' cross-channel behaviors, enterprises can more effectively target customers and improve the accuracy and efficiency of marketing activities.

[0019] Optionally, determining the customer's consultation intention according to the matching result includes: Determining whether the order information is consistent with the product information according to the matching result; If they are consistent, the order information is parsed to determine the order purchase time; Determine whether there is any suspicious after-sales service based on the purchase time of the order; If so, the suspicious after-sales service will be determined as the customer's consultation intention.

[0020] Through this solution, by quickly identifying customers' consultation intentions, enterprises can respond to customer needs more effectively and improve customer service quality and satisfaction. By verifying the consistency of order information and product information, misunderstandings and customer complaints caused by information errors are reduced. By promptly identifying and handling suspicious after-sales situations, fraud risks are reduced. By analyzing customers' consultation intentions, enterprises can better understand market needs and customer pain points.

[0021] Optionally, before responding to the customer according to the preset response logic and the response ideas, the method further includes: Analyze the response ideas to determine if there are any sensitive issues; If so, determine the feasibility of the response according to the preset response logic; The feasibility of the reply is compared with the preset feasibility. If the feasibility of the reply is lower than the preset feasibility, the customer consultation intention, the customer basic information and the customer communication content are sent to the manual customer service end.

[0022] This solution can significantly improve customer service response speed and reduce customer waiting time by automatically analyzing customer inquiries and generating preliminary responses. Automatically generated responses can be based on preset logic and standards to ensure that the information received by customers is accurate and consistent, thereby improving overall service quality. Automatic processing of simple and repetitive customer inquiries can reduce the need for manual customer service, thereby reducing operating costs. Fast and accurate responses can enhance customer trust and satisfaction with the brand.

[0023] In a second aspect, the present application provides a cross-channel customer interaction management system based on data fusion, the system comprising: Demand analysis module, used to obtain customer communication needs, analyze the customer communication needs, and determine customer basic information and customer communication content; An intention determination module is used to determine the customer's consultation intention based on multi-channel data, according to the customer's basic information and the customer's communication content; The reply module is used to determine the reply ideas according to the customer's consultation intention, and respond to the customer according to the preset reply logic and the reply ideas.

[0024] Optionally, when the intention determination module determines the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, it is used to: Analyze the customer communication content to determine whether the customer communication content contains product information; If included, determining whether the customer has channel information based on the multi-channel data and the customer basic information; If there is channel information, determining the order information of the customer according to the channel information; Matching the product information with the order information; Based on the matching results, the customer's consultation intention is determined.

[0025] Optionally, when the intention determination module determines the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, it is used to: If the channel information does not include order information, determining the associated user based on the product information and the multi-channel data; Analyze the order records of the associated users to determine the reasons for the transactions; Determine the customer's consultation intention based on the reasons for the transaction and the customer's consultation intention.

[0026] Optionally, when the intention determination module determines the customer's consultation intention according to the transaction reason and the customer's consultation intention, it is used to: According to the transaction reason, retrieve the communication response record of the associated user; Based on the communication response records, retrieve the transaction time and operation records before the transaction; Determining whether the communication response record has a positive impact on the transaction based on the operation record and the transaction time; If so, the customer's consultation intention is determined based on the communication response record and the customer's consultation intention.

[0027] Optionally, when the intention determination module determines the associated user according to the product information and the multi-channel data, it is used to: According to the product information, the multi-channel data is retrieved to determine whether there is a user who inquires about the product corresponding to the product information or purchases the product corresponding to the product information; If so, retrieve relevant information of the user who inquired about the product corresponding to the product information or purchased the product corresponding to the product information; Retrieving the purchase records of the multi-channel data according to the product information; Analyze the purchase records to determine important attributes of the purchaser; The relevant information is filtered according to the important attributes, and the associated users are determined according to the filtering results.

[0028] Optionally, when the intention determination module determines the customer's consultation intention based on the multi-channel data, according to the customer basic information and the customer communication content, it is used to: If channel information exists, determining historical communication information of the customer based on the multi-channel data; Comparing the historical communication information with the customer communication content to determine whether the historical communication information contains related content that is consistent with the customer communication content; If so, retrieve and analyze the response record of the related content, and determine the direction of the customer's doubts based on the analysis results of the response record; According to the direction of the doubt, the customer's consultation intention is determined.

[0029] Optionally, when the intention determination module determines whether the customer has channel information based on the customer basic information, it is used to: Determine the identity information and login information of the customer based on the basic customer information; Matching the identity information and the login information with the multi-channel data respectively, and determining whether there is completely consistent channel data according to the matching results; If yes, it is determined that the customer has channel information; If not, analyzing the customer's communication content, and determining the customer's communication habits based on the analysis results; Analyzing whether there is channel data in the multi-channel data whose similarity to the communication habit reaches a preset similarity; If yes, the channel data, the identity information and the login information are analyzed to determine whether the customer has channel information.

[0030] Optionally, when the intention determination module determines the customer's consultation intention according to the matching result, it is used to: Determining whether the order information is consistent with the product information according to the matching result; If they are consistent, the order information is parsed to determine the order purchase time; Determine whether there is any suspicious after-sales service based on the purchase time of the order; If so, the suspicious after-sales service will be determined as the customer's consultation intention.

[0031] Optionally, the cross-channel customer interaction management system further includes an information sending module, which is used to: Analyze the response ideas to determine if there are any sensitive issues; If so, determine the feasibility of the response according to the preset response logic; The feasibility of the reply is compared with the preset feasibility. If the feasibility of the reply is lower than the preset feasibility, the customer consultation intention, the customer basic information and the customer communication content are sent to the manual customer service end. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0033] Figure 1 A schematic diagram of an application scenario provided for an embodiment of the present application; Figure 2 A flowchart of a cross-channel customer interaction management method based on data fusion provided in one embodiment of the present application; Figure 3 A schematic diagram of the structure of a cross-channel customer interaction management system based on data fusion provided in one embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0036] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0037] Traditional customer service and interaction management usually rely on a single communication channel and isolated customer information, making it difficult for companies to fully understand customer needs and behaviors. Therefore, how to effectively use these multi-channel data to improve customer service quality and increase customer satisfaction has become the focus of companies.

[0038] Based on this, the present application provides a cross-channel customer interaction management system and method based on data fusion, which obtains customer communication needs, analyzes the customer communication needs, determines customer basic information and customer communication content; based on multi-channel data, determines customer consultation intention according to the customer basic information and the customer communication content; determines the reply ideas according to the customer consultation intention, and responds to the customer according to the preset reply logic and the reply ideas. Through automated processing and intelligent response, the burden of manual customer service is reduced and the efficiency of handling customer consultation is improved. Real-time data analysis enables the system to respond to user needs quickly and reduces customer waiting time. Personalized services are based on customers' multi-channel data and communication history, and can provide services that are more in line with customer needs and improve customer satisfaction. Consistent service experience, no matter which channel customers communicate through, they can get consistent service standards. Integrate multi-channel data to provide customers with comprehensive analysis reports to help companies better understand customer behavior. Using machine learning and natural language processing technology, the system can more accurately identify customers' consultation intentions and reduce misunderstandings and repeated consultations. Improve the accuracy of consultation responses to ensure that customers get the right information and suggestions.

[0039] Figure 1 A schematic diagram of an application scenario provided for this application, in which the method provided by this application is applied. Specifically, the method provided by this application is applied to any server, and the server interacts with the channel application, and can instantly capture the communication needs of customers on the channel application, so that the enterprise can quickly respond to customer problems. Through multi-channel data collection, more comprehensive customer information can be obtained to improve service quality. By analyzing the customer communication content on the channel application, you can deeply understand the customer's intentions and needs, so as to provide more personalized services. It can identify the customer's emotions and attitudes, which helps companies to better adjust their service strategies. Through customer basic information, companies can more accurately identify customers and provide customized services for each customer. Integrating customer communication content from different channels can build a more comprehensive customer portrait. Determining customer consultation intentions helps companies provide services based on the specific needs of customers and improve service efficiency. By identifying the main consultation intentions, companies can reasonably allocate resources and focus on the most critical service areas. The determination of the reply ideas ensures the strategic and forward-looking nature of the service and helps to improve customer satisfaction. Replying to customers on the channel application through pre-set reply logic can ensure the consistency and professionalism of the service. The automated reply process can greatly improve the speed of handling consultations and reduce customer waiting time. Preset response logic ensures accuracy and consistency of responses, reducing errors and omissions.

[0040] For specific implementation methods, please refer to the following embodiments.

[0041] Figure 2This is a flow chart of a cross-channel customer interaction management method based on data fusion provided in an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201. Obtain customer communication needs, analyze customer communication needs, and determine customer basic information and customer communication content; Customer communication needs can be the problems that customers hope to solve, the information they hope to obtain, or the goals they hope to achieve through communication when interacting with the company; such as consulting product information, after-sales service, complaints and suggestions, technical support, etc.

[0042] Customer basic information can be the customer's basic information, such as name, gender, age, contact information, residential address, occupation, income level, purchase history, preferences, etc.

[0043] Customer communication content can be the specific information expressed by customers during the communication process with the company, such as text, voice, image, video and other forms of content.

[0044] Specifically, first, collect customer communication records through various communication channels (telephone, email, instant messaging, social media, website, etc.), including data in various forms such as text, voice, and image. Use natural language processing (NLP) technology to pre-process the communication content, such as word segmentation, part-of-speech tagging, entity recognition, etc., in order to extract key information. Conduct in-depth analysis of the pre-processed communication content, including sentiment analysis, topic modeling, keyword extraction, etc., to understand the customer's explicit needs and potential emotions. Combine the customer's historical communication records to analyze the customer's behavior patterns and further explore the customer's needs. Extract the customer's basic personal information from multi-channel data, such as name, contact information, purchase history, etc. Classify and label the communication content to quickly identify the customer's communication purpose.

[0045] S202. Determine the customer's consultation intention based on multi-channel data, basic customer information and customer communication content; Multichannel data can be customer data from different communication channels such as website, social media, mobile app, email, phone, chat platform, etc.

[0046] Customer consultation intention can be the specific needs or problems expressed by the customer during communication. By analyzing the customer's communication content and other relevant information, the customer's intentions and expectations can be inferred.

[0047] Specifically, we use data mining and machine learning algorithms to identify customers’ consulting intentions by combining their basic information and communication content. We analyze customers’ behavior trajectories on multiple channels and build customer portraits to help determine the accuracy of their consulting intentions.

[0048] S203. Determine the answering ideas based on the customer's inquiry intention, and respond to the customer based on the preset answering logic and answering ideas.

[0049] The response idea can be the ideas and methods proposed by customer service personnel or the system to deal with customer problems after understanding and analyzing the customer's inquiry intention.

[0050] Canned response logic can be a set of predefined response rules and guidelines that guide the system or human customer service on how to respond when handling customer inquiries.

[0051] Specifically, design the corresponding response framework and ideas based on the identified customer consultation intentions. Use the preset response logic, such as decision trees, rule engines, etc., to generate preliminary response content. Review the generated response content to ensure that it complies with the company's service standards and policies. Send the response content to the customer through the corresponding communication channel to complete the interaction.

[0052] Through this solution, the burden of manual customer service is reduced and the efficiency of handling customer inquiries is improved through automated processing and intelligent response. Real-time data analysis enables the system to respond quickly to user needs and reduce customer waiting time. Personalized services are based on customers' multi-channel data and communication history, which can provide services that are more in line with customer needs and improve customer satisfaction. A consistent service experience means that customers can receive consistent service standards regardless of which channel they communicate through. Integrate multi-channel data to provide customers with comprehensive analysis reports to help companies better understand customer behavior. Using machine learning and natural language processing technology, the system can more accurately identify customers' consultation intentions and reduce misunderstandings and repeated consultations. Improve the accuracy of consultation responses to ensure that customers receive the correct information and advice.

[0053] In some embodiments, the customer communication content is analyzed to determine whether the customer communication content contains product information; if so, based on multi-channel data and according to the customer's basic information, it is determined whether the customer has channel information; if channel information exists, the customer's order information is determined based on the channel information; the product information is matched with the order information; and based on the matching results, the customer's inquiry intention is determined.

[0054] Product information can be detailed information about a specific product, such as the product's name, brand, model, price, specifications, functions, size, color, weight, production date, shelf life, manufacturer, supplier, etc.

[0055] Channel information can be relevant data on various channels of communication and transactions between an enterprise and its customers.

[0056] Order information may be specific transaction data submitted by a customer when purchasing goods or services.

[0057] Specifically, collect customer communication records from various communication channels, including data in various forms such as text, voice, and images. Integrate this data into a unified data platform for subsequent analysis and processing. Preprocess the collected communication content, including word segmentation, part-of-speech tagging, and entity recognition. Use NLP technology to analyze the communication content and extract key information, such as product name, question type, and customer sentiment. Analyze the customer communication content to determine whether it contains product information, such as model, brand, color, etc. If it contains product information, extract this information for subsequent matching with order information. Extract the customer's basic personal information from multi-channel data based on the contact information or other identity identifiers provided by the customer. This information may include name, contact information, purchase record, service record, etc. Combined with the customer's basic information, determine whether the customer has active interaction records on one or more channels. This channel information may include website visit records, social media interactions, phone call records, etc. Extract the customer's order information from multi-channel data based on the customer's basic information and channel information. Order information may include purchase date, product details, payment method, delivery information, etc. Match the product information extracted from the customer's communication content with the order information. Confirm whether the product the customer is inquiring about is the same as the product he / she purchased, or whether it is a related product. Based on the matching results of product information and order information, combined with the customer's communication content, determine the customer's inquiry intention. For example, if the match is successful, the customer may need after-sales service or product consultation; if the match is unsuccessful, the customer may be inquiring about new products or related products.

[0058] Through this solution, the burden of manual customer service is reduced and the efficiency of handling customer inquiries is improved through automated processing and intelligent response. Customers are automatically classified according to their behavior and communication content to facilitate targeted services. Real-time data analysis enables the system to respond quickly to user needs and reduce customer waiting time. Personalized services are based on customers' multi-channel data and communication history, and can provide services that are more in line with customer needs and improve customer satisfaction. Through historical data, predict possible customer needs and prepare in advance. Provide personalized product or service recommendations based on customer preferences and purchase history. Consistent service experience, no matter which channel customers communicate through, they can get consistent service standards. Integrate multi-channel data to provide customers with comprehensive analysis reports to help companies better understand customer behavior and market trends.

[0059] In some embodiments, if the channel information does not include order information, the associated users are determined based on the product information and multi-channel data; the order records of the associated users are analyzed to determine the reasons for the transactions; and the customer's consultation intention is determined based on the reasons for the transactions and the content of the customer communication.

[0060] Order information can be a record created when a customer purchases goods or services, containing all the details related to the order.

[0061] Related users can be a group of users who have a direct or indirect relationship with a specific product.

[0062] The order record may be a detailed record of the specific transaction activities for each order.

[0063] The reason for the transaction can be the factor that leads the customer to finally complete the purchase decision.

[0064] Specifically, collect customer data from different channels, including social media, email, instant messaging, phone recordings, website behavior logs, etc. Integrate data from different channels to ensure data consistency and analyzability. This may involve data cleaning, conversion, and merging. Extract basic customer information from the integrated data, such as name, address, contact information, purchase history, etc. Analyze product information, including product description, price, specifications, inventory status, sales records, etc. Use product information and customer basic information to identify user groups related to specific products through algorithms. Analyze the order records of associated users, identify purchase behavior patterns, and determine the reasons for transactions. This may include analyzing purchase frequency, purchase amount, purchase time, purchase channel, etc. Based on the results of order record analysis, determine the main factors that affect customer purchase decisions, such as price, promotional activities, brand reputation, etc. Analyze customer communication content, including questions, feedback, complaints, etc., to understand the specific needs and problems of customers. Combine the analysis of transaction reasons and customer communication content, and use natural language processing and machine learning technology to identify customers' consultation intentions.

[0065] Through this solution, customized product recommendations and services are provided by analyzing customer historical data and communication content. Real-time analysis of customer communication provides immediate answers and help. Predict problems that customers may encounter and prepare in advance. Improve sales conversion rate through accurate consultation intention identification. Form a unified customer view by integrating data from different channels. Use machine learning algorithms to predict customer behavior and identify consultation intentions. Automated customer service is achieved through chatbots, automatic replies and other functions.

[0066] In some embodiments, based on the reason for the transaction, the communication response record of the associated user is retrieved; based on the communication response record, the transaction time and the operation record before the transaction are retrieved; based on the operation record and the transaction time, it is determined whether the communication response record has a positive impact on the transaction; if so, the customer's consultation intention is determined based on the communication response record and the customer's communication content.

[0067] Communication response records can be records of a company's responses to customer inquiries, complaints, feedback, etc. during the process of interaction with customers.

[0068] The transaction time can be the time when the customer completes the purchase transaction. This is usually a specific date and time, recording the entire process from the customer browsing the product to the payment completion.

[0069] Operation records can be records of various operations performed by customers during the purchase process.

[0070] Positive influence can be a behavior or event that has a positive effect or promotes the outcome.

[0071] Specifically, collect customer data from multiple channels, including order information, communication records, behavioral data, etc. Use data integration tools to merge data from different sources into a unified database. Identify users related to specific transactions based on product information and order data. Ensure consistency of user identity across multiple channels. Retrieve the communication response records of associated users from the integrated database. Including emails, chat records, social media interactions, etc. Retrieve the transaction time records of associated users. Obtain the user's operation records before the transaction, such as page browsing, click streams, search history, etc. Analyze the user's operation records before the transaction to determine the relationship between user behavior and the transaction. Use time series analysis or user behavior analysis technology to identify user behavior patterns. Evaluate the impact of communication response records on the transaction and determine whether there is a positive impact. This may involve sentiment analysis, text analysis, or comparative analysis. If the communication response records have a positive impact on the transaction, further analyze these communication contents. Use natural language processing (NLP) technology to identify customer consultation intentions.

[0072] This solution can quickly identify and respond to customer inquiries, reducing waiting time. Provide personalized product or service recommendations based on customer historical behavior and inquiry content. Improve problem-solving efficiency by accurately identifying customer inquiry intentions. High-quality service experience helps improve customer retention. Use customer inquiry intention data to optimize marketing activities and increase conversion rates. Reduce the operating costs of the customer service department through automated processing.

[0073] In some embodiments, based on the product information, multi-channel data is retrieved to determine whether there is a user who has inquired about the product corresponding to the product information or purchased the product corresponding to the product information; if so, the relevant information of the user who has inquired about the product corresponding to the product information or purchased the product corresponding to the product information is retrieved; based on the product information, the purchase records of multi-channel data are retrieved; the purchase records are analyzed to determine the important attributes of the purchaser; based on the important attributes, relevant information is filtered, and based on the filtering results, the associated users are determined.

[0074] The product corresponding to the consultation product information may be the specific product mentioned by the customer during the consultation process.

[0075] The product corresponding to the purchased product information may be the product actually purchased by the customer.

[0076] Important attributes can be features that have a significant impact on customer purchasing behavior.

[0077] Specifically, collect customer data from various channels, including but not limited to social media, email, instant messaging, phone records, website behavior, etc. Use data integration tools to merge data scattered across different channels into a unified database to ensure data consistency and analyzability. Extract information related to specific products from the database, such as product name, description, price, inventory, etc. Analyze multi-channel data to identify users who have consulted or purchased specific products. This may involve screening mentions on social media, search records on the website, customer service records, etc. For identified users, retrieve their relevant information, including but not limited to user ID, contact information, purchase history, consultation records, etc. Analyze users' purchase records, including purchase time, purchase frequency, purchased product type, purchase amount, etc. This may involve technologies such as time series analysis and customer behavior analysis. Extract important attributes from purchase records, such as purchase frequency, purchase amount, purchased product type, etc., which may have an important impact on the selection of associated users. Based on the determined important attributes, screen out potential associated users with similar purchase behaviors. This may involve technologies such as cluster analysis and association rule mining. Comprehensively analyze users' consultation records, purchase behaviors, important attributes and other information to ultimately determine associated users related to specific products.

[0078] Through this solution, sales conversion rates can be improved through precision marketing for related users. By analyzing multi-channel data, customers who are interested in or have already purchased a specific product can be more accurately targeted. Quick response and personalized services for potential customers can shorten the sales cycle. Through multi-channel data, customer interaction patterns can be better understood and customer interaction can be enhanced. Based on purchase record analysis, inventory management can be optimized to reduce inventory backlogs. By analyzing customer needs, existing products can be improved or new products can be developed. For identified related users, precision marketing can be more effectively carried out to improve the conversion rate of marketing activities.

[0079] In some embodiments, if channel information exists, the customer's historical communication information is determined based on multi-channel data; the historical communication information is compared with the customer's communication content to determine whether the historical communication information contains related content that is consistent with the customer's communication content; if so, the response records of the related content are retrieved and analyzed, and the direction of the customer's doubts is determined based on the analysis results of the response records; based on the direction of the doubts, the customer's consultation intention is determined.

[0080] Historical communication information can be all past communication records between customers and businesses, such as phone call records, emails, instant message conversations, social media interactions, website messages, etc.

[0081] Related content can be the parts of the customer's historical communication information that match or are related to the current inquiry content. For example, if a customer has previously asked about the product warranty policy and now asks a similar question, the previous inquiry content about the warranty policy can be considered related content.

[0082] Response records can be records of the company's responses or solutions to historical customer communications, such as customer service staff's responses, technical support suggestions, and after-sales service processing results.

[0083] Doubts can be areas of doubt or uncertainty expressed by customers during communication.

[0084] Specifically, if channel information exists, collect customer data from different communication channels (telephone, email, instant messaging, social media, website, etc.). Clean and standardize the collected data to remove noise and inconsistent information. Retrieve the customer's historical communication records from the integrated data. Classify the historical records according to communication content, time, channel, etc. Analyze the communication content of the current customer using natural language processing (NLP) technology. Identify keywords and phrases in the communication content to understand the customer's needs and intentions. Apply matching algorithms to compare historical communication information with current communication content to find similar or related expressions. Determine whether there is related content in the historical communication information that is consistent with the current communication content. If related content is found, retrieve the corresponding response records. Analyze the response records to understand the company's previous solutions or service provision. Based on the analysis results of the response records, identify possible questions or problems that the customer may have. Summarize the direction of the customer's doubts, such as product quality, after-sales service, price issues, etc. According to the direction of doubt, classify the customer's consultation intention, such as product consultation, technical support, complaint suggestions, etc. Verify the customer's consultation intention through further communication or data analysis.

[0085] Through this solution, by identifying related users, enterprises can avoid repeatedly answering the same questions and reduce customer service workload. Analyzing customer communication history can help enterprises optimize service processes and reduce the occurrence of common problems. Personalized services and quick responses can improve customer loyalty, reduce customer churn, and provide high-quality customer service.

[0086] In some embodiments, the customer's identity information and login information are determined based on the customer's basic information; the identity information and login information are matched with multi-channel data respectively, and based on the matching results, it is determined whether there is completely consistent channel data; if so, it is determined that the customer has channel information; if not, the customer's communication content is analyzed, and based on the analysis results, the customer's communication habits are determined; it is analyzed in the multi-channel data whether there is channel data whose similarity to the communication habits reaches a preset similarity; if so, the channel data, identity information and login information are analyzed to determine whether the customer has channel information.

[0087] Identity information can be personal information that can uniquely identify a customer, such as name, ID number, passport number, date of birth, gender, etc.

[0088] The login information may be information used by a user to log into the system, such as a user name, password, verification code, etc.

[0089] Channel data can be data collected through different communication channels, which may include but are not limited to email, social media, telephone, website, mobile application, etc.; channel data includes user behavior data, transaction data, communication content, etc.

[0090] Channel information can be activity information on a specific channel, such as customers’ likes, comments, and sharing behaviors on social media, or purchase records and browsing history on e-commerce websites.

[0091] Customer communication content can be in the form of text, pictures, audio or video generated when customers communicate with the company.

[0092] Communication habits can be specific behavioral patterns that customers exhibit when communicating, such as using specific vocabulary, the way they ask questions, and their time preferences for communication.

[0093] Communication habit similarity can be calculated by comparing the communication habits of two or more customers and calculating the degree of similarity between them.

[0094] Specifically, collect basic information such as the customer's name, contact information (telephone number, email, etc.), username, password, etc. from channels such as CRM systems, order databases, and registration forms. Match the collected customer identity information (such as name, contact information) with multi-channel data (such as social media, shopping websites, forums, etc.). Match the customer's login information (such as username, password) with the login information in the multi-channel data. Analyze the matching results to see if there is completely consistent channel data. This includes checking whether the identity information and login information have corresponding records in multiple channels. If the identity information and login information are completely consistent with the multi-channel data, it can be determined that the customer has channel information. If there is no completely consistent matching result, proceed to the next step. If there is no completely consistent matching result, analyze the customer's communication content in multiple channels, including language style, common vocabulary, questioning methods, etc. Analyze the customer's communication habits, such as whether they tend to use specific expressions, whether they often ask certain types of questions, etc. Compare the analyzed customer communication habits with the communication habits in the multi-channel data and calculate the similarity. Set a preset similarity threshold to determine the similarity of communication habits. If the similarity of communication habits reaches the preset threshold, the channel data, identity information and login information are further analyzed to determine whether the customer has channel information. For example, a comprehensive analysis of the customer's activity patterns, purchasing behavior, interaction frequency, etc. in multiple channels is conducted. Based on the analysis results, it is confirmed whether the customer has channel information, and the analysis process and conclusions are recorded, and the analysis results are updated in the customer's file.

[0095] Through this solution, through the matching and analysis of multi-channel data, enterprises can more accurately identify and associate multiple accounts and interactions of the same customer, thereby improving the integrity of customer identification. By comprehensively analyzing multi-channel data, enterprises can gain a deeper understanding of customers' purchasing behaviors, preferences, and needs, providing a basis for personalized marketing and services. Determining whether channel information exists for customers helps provide a consistent service experience, even if customers interact on different channels. By identifying customers' cross-channel behaviors, enterprises can more effectively target customers and improve the accuracy and efficiency of marketing activities.

[0096] In some embodiments, based on the matching results, it is determined whether the order information is consistent with the product information; if they are consistent, the order information is parsed to determine the order purchase time; based on the order purchase time, it is determined whether there is suspicious after-sales service; if so, the suspicious after-sales service is determined as a customer inquiry intention.

[0097] The purchase time can be the specific time point when the customer completes the purchase transaction.

[0098] Suspicious after-sales service requests may be those that may indicate fraud, product quality issues, or other abnormal situations.

[0099] Specifically, match the order information with the product information to verify whether the product information in the order is accurate. Compare the order information and product information to check whether the product name, price, inventory and other information are consistent. If any inconsistencies are found, record them and further investigate the reasons. For orders that match consistently, parse the order information to determine the purchase time of the order. Record the order purchase time for subsequent analysis. Query the after-sales database to check whether there are after-sales requests near the order purchase time. Identify after-sales requests that are close to the purchase time, which may indicate suspicious after-sales situations. Analyze suspicious after-sales requests to determine the questions or needs that the customer may inquire about. Based on the analysis results, determine the suspicious after-sales as the customer's specific consultation intention, such as product quality issues, after-sales service needs, refund requests, etc.

[0100] Through this solution, by quickly identifying customers' consultation intentions, enterprises can respond to customer needs more effectively and improve customer service quality and satisfaction. By verifying the consistency of order information and product information, misunderstandings and customer complaints caused by information errors are reduced. By promptly identifying and handling suspicious after-sales situations, fraud risks are reduced. By analyzing customers' consultation intentions, enterprises can better understand market needs and customer pain points.

[0101] In some embodiments, before responding to customers according to preset response logic and response ideas, it also includes: analyzing the response ideas to determine whether there are sensitive issues; if so, determining the feasibility of the response according to the preset response logic; comparing the feasibility of the response with the preset feasibility. If the feasibility of the response is lower than the preset feasibility, the customer's consultation intention, customer basic information and customer communication content are sent to the manual customer service end.

[0102] Sensitive issues may include those that may involve personal privacy, security issues or other sensitive information.

[0103] Canned reply logic can be pre-defined rules and algorithms that guide the behavior of the automatic reply system.

[0104] Response feasibility can be the applicability and effectiveness of automatically generated responses in specific situations, such as whether the response is accurate, whether it can meet customer needs, and whether it complies with corporate policies.

[0105] Preset feasibility can be the standards set by the enterprise regarding the feasibility of automatic replies, such as accuracy, compliance, customer satisfaction, etc.

[0106] The manual customer service terminal may be an interface or system used for manual customer service operations.

[0107] Specifically, receive customer consultation content from customer service channels (such as applications, APPs, e-commerce platforms, emails, chatbots, etc.). Analyze the consultation content, including keyword extraction, sentiment analysis, etc., to understand the customer's questions and intentions. Use a pre-set list of sensitive words or a sensitive question library to identify whether the customer consultation contains sensitive questions. Analyze the context of the consultation content to determine whether there are sensitive or complex questions. Match the customer consultation content with the preset response logic to determine whether there are ready-made responses that can be applied. According to the preset response logic, evaluate whether the automatically generated response is feasible and whether it complies with the company's service standards and policies. Determine the preset response feasibility standards, which may include accuracy, completeness, compliance, etc. Compare the automatically generated response with the preset feasibility standards to evaluate whether it meets the requirements. If the feasibility of the automatically generated response is lower than the preset standard, mark it as requiring manual intervention. Send the customer's consultation intention, basic customer information, and customer communication content to the manual customer service end so that the manual customer service can provide more accurate and personalized services.

[0108] This solution can significantly improve customer service response speed and reduce customer waiting time by automatically analyzing customer inquiries and generating preliminary responses. Automatically generated responses can be based on preset logic and standards to ensure that the information received by customers is accurate and consistent, thereby improving overall service quality. Automatic processing of simple and repetitive customer inquiries can reduce the need for manual customer service, thereby reducing operating costs. Fast and accurate responses can enhance customer trust and satisfaction with the brand.

[0109] Figure 3 A schematic diagram of a cross-channel customer interaction management system based on data fusion is provided in one embodiment of the present application, such as Figure 3 As shown, the cross-channel customer interaction management system 300 based on data fusion of this embodiment includes: a demand analysis module 301 , an intention determination module 302 , and a reply module 303 .

[0110] Demand analysis module 301, used to obtain customer communication needs, analyze the customer communication needs, and determine customer basic information and customer communication content; The intention determination module 302 is used to determine the customer's consultation intention based on multi-channel data, according to the customer basic information and the customer communication content; The reply module 303 is used to determine a reply idea according to the customer's consultation intention, and respond to the customer according to the preset reply logic and the reply idea.

[0111] Optionally, when determining the customer's consultation intention based on the multi-channel data, the customer basic information and the customer communication content, the intention determination module 302 is used to: Analyze the customer communication content to determine whether the customer communication content contains product information; If included, determining whether the customer has channel information based on the multi-channel data and the customer basic information; If there is channel information, determining the order information of the customer according to the channel information; Matching the product information with the order information; Based on the matching results, the customer's consultation intention is determined.

[0112] Optionally, when determining the customer's consultation intention based on the multi-channel data, the customer basic information and the customer communication content, the intention determination module 302 is used to: If the channel information does not include order information, determining the associated user based on the product information and the multi-channel data; Analyze the order records of the associated users to determine the reasons for the transactions; Determine the customer's consultation intention based on the reasons for the transaction and the customer's consultation intention.

[0113] Optionally, when determining the customer's consultation intention according to the transaction reason and the customer's consultation intention, the intention determination module 302 is used to: According to the transaction reason, retrieve the communication response record of the associated user; Based on the communication response records, retrieve the transaction time and operation records before the transaction; Determining whether the communication response record has a positive impact on the transaction based on the operation record and the transaction time; If so, the customer's consultation intention is determined based on the communication response record and the customer communication content.

[0114] Optionally, when determining the associated user according to the product information and the multi-channel data, the intention determination module 302 is used to: According to the product information, the multi-channel data is retrieved to determine whether there is a user who inquires about the product corresponding to the product information or purchases the product corresponding to the product information; If so, retrieve relevant information of the user who inquired about the product corresponding to the product information or purchased the product corresponding to the product information; Retrieving the purchase records of the multi-channel data according to the product information; Analyze the purchase records to determine important attributes of the purchaser; The relevant information is filtered according to the important attributes, and the associated users are determined according to the filtering results.

[0115] Optionally, when determining the customer's consultation intention based on the multi-channel data, the customer basic information and the customer communication content, the intention determination module 302 is used to: If channel information exists, determining historical communication information of the customer based on the multi-channel data; Comparing the historical communication information with the customer communication content to determine whether the historical communication information contains related content consistent with the customer communication content; If so, retrieve and analyze the response record of the related content, and determine the direction of the customer's doubts based on the analysis results of the response record; According to the direction of the doubt, the customer's consultation intention is determined.

[0116] Optionally, when the intention determination module 302 determines whether the customer has channel information according to the customer basic information, it is used to: Determine the identity information and login information of the customer based on the basic customer information; Matching the identity information and the login information with the multi-channel data respectively, and determining whether there is completely consistent channel data according to the matching results; If yes, it is determined that the customer has channel information; If not, analyzing the customer's communication content, and determining the customer's communication habits based on the analysis results; Analyzing whether there is channel data in the multi-channel data whose similarity to the communication habit reaches a preset similarity; If yes, the channel data, the identity information and the login information are analyzed to determine whether the customer has channel information.

[0117] Optionally, when determining the customer's consultation intention according to the matching result, the intention determination module 302 is used to: Determining whether the order information is consistent with the product information according to the matching result; If they are consistent, the order information is parsed to determine the order purchase time; Determine whether there is any suspicious after-sales service based on the purchase time of the order; If so, the suspicious after-sales service will be determined as the customer's consultation intention.

[0118] Optionally, the cross-channel customer interaction management system 300 further includes an information sending module 304, which is used to: Analyze the response ideas to determine if there are any sensitive issues; If so, determine the feasibility of the response according to the preset response logic; The feasibility of the reply is compared with the preset feasibility. If the feasibility of the reply is lower than the preset feasibility, the customer consultation intention, the customer basic information and the customer communication content are sent to the manual customer service end.

[0119] The system of this embodiment can be used to execute the method of any of the above embodiments. The implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A cross-channel customer interaction management method based on data fusion, characterized in that: include: Obtain customer communication needs, analyze said customer communication needs, and determine customer basic information and customer communication content; Based on multi-channel data, determine the customer's consultation intention according to the customer's basic information and the customer's communication content; Determine the response ideas based on the customer's inquiry intention, and respond to the customer based on the preset response logic and the response ideas.

2. The method according to claim 1, characterized in that: The multi-channel data is used to determine the customer's consultation intention according to the customer's basic information and the customer's communication content, including: Analyze the customer communication content to determine whether the customer communication content contains product information; If included, determining whether the customer has channel information based on the multi-channel data and the customer basic information; If there is channel information, determining the order information of the customer according to the channel information; Matching the product information with the order information; Based on the matching results, the customer's consultation intention is determined.

3. The method according to claim 2, characterized in that The multi-channel data is used to determine the customer's consultation intention according to the customer's basic information and the customer's communication content, including: If the channel information does not include order information, determining the associated user based on the product information and the multi-channel data; Analyze the order records of the associated users to determine the reasons for the transactions; Determine the customer's consultation intention based on the reasons for the transaction and the content of the customer communication.

4. The method according to claim 3, characterized in that: Determining the customer's consultation intention based on the transaction reason and the customer's communication content includes: According to the transaction reason, retrieve the communication response record of the associated user; Based on the communication response records, retrieve the transaction time and operation records before the transaction; Determining whether the communication response record has a positive impact on the transaction based on the operation record and the transaction time; If so, the customer's consultation intention is determined based on the communication response record and the customer communication content.

5. The method according to claim 3, characterized in that: The determining of associated users according to the product information and the multi-channel data includes: According to the product information, the multi-channel data is retrieved to determine whether there is a user who inquires about the product corresponding to the product information or purchases the product corresponding to the product information; If so, retrieve relevant information of the user who inquired about the product corresponding to the product information or purchased the product corresponding to the product information; Retrieving the purchase records of the multi-channel data according to the product information; Analyze the purchase records to determine important attributes of the purchaser; The relevant information is filtered according to the important attributes, and the associated users are determined according to the filtering results.

6. The method according to claim 2, characterized in that The multi-channel data is used to determine the customer's consultation intention according to the customer's basic information and the customer's communication content, including: If channel information exists, determining historical communication information of the customer based on the multi-channel data; Comparing the historical communication information with the customer communication content to determine whether the historical communication information contains related content that is consistent with the customer communication content; If so, retrieve and analyze the response record of the related content, and determine the direction of the customer's doubts based on the analysis results of the response record; According to the direction of the doubt, the customer's consultation intention is determined.

7. The method according to claim 2, characterized in that: Determining whether the customer has channel information according to the customer basic information includes: Determine the identity information and login information of the customer based on the basic customer information; Matching the identity information and the login information with the multi-channel data respectively, and determining whether there is completely consistent channel data according to the matching results; If yes, it is determined that the customer has channel information; If not, analyzing the customer's communication content, and determining the customer's communication habits based on the analysis results; Analyzing whether there is channel data in the multi-channel data whose similarity to the communication habit reaches a preset similarity; If yes, the channel data, the identity information and the login information are analyzed to determine whether the customer has channel information.

8. The method according to claim 2, characterized in that: Determining the customer's consultation intention according to the matching result includes: Determining whether the order information is consistent with the product information according to the matching result; If they are consistent, the order information is parsed to determine the order purchase time; Determine whether there is any suspicious after-sales service based on the purchase time of the order; If so, the suspicious after-sales service will be determined as the customer's consultation intention.

9. The method according to claim 1, characterized in that: Before responding to the customer according to the preset response logic and the response ideas, the method further includes: Analyze the response ideas to determine if there are any sensitive issues; If so, determine the feasibility of the response according to the preset response logic; The feasibility of the reply is compared with the preset feasibility. If the feasibility of the reply is lower than the preset feasibility, the customer consultation intention, the customer basic information and the customer communication content are sent to the manual customer service end.

10. A cross-channel customer interaction management system based on data fusion, characterized in that: include: Demand analysis module, used to obtain customer communication needs, analyze the customer communication needs, and determine customer basic information and customer communication content; An intention determination module is used to determine the customer's consultation intention based on multi-channel data, according to the customer's basic information and the customer's communication content; The reply module is used to determine the reply ideas according to the customer's consultation intention, and respond to the customer according to the preset reply logic and the reply ideas.

Images