Intelligent user service platform based on big data

Abstract

The invention discloses an intelligent user service platform based on big data, which comprises a service data analysis management module, a server semantic analysis module and an intelligent service interaction feedback module, the operation method of the platform comprises the following steps: step 1, integrating data, processes, resources, organizations and functions related to business products in different life cycles based on enterprise business services to a unified intelligent user service platform for analysis and management of service data; 2, on the basis of shallow-layer user consultation text semantic analysis, keywords are marked according to semantic similarity and a vector model, and the similarity between the keywords is calculated; step 3, constructing a rule of progressive service of the intelligent user server to the user level; 4, performing question semantic analysis and information retrieval on the questions consulted by the user to form final intelligent service interaction feedback; the problem of invalid service of intelligent customer service under complex conditions is solved, and the user service satisfaction degree and the user intelligent service quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise platform intelligent user services, and in particular to an intelligent user service platform based on big data. Background Art

[0002] In the existing business environment, all enterprise systems are equipped with intelligent user service platforms. The introduction of intelligent customer service in the platform has greatly reduced the workload of employees and saved the company's manpower expenditure. However, judging from the data on the dissatisfaction rate of intelligent services, users evaluate that the intelligent customer service has poor problem-solving capabilities and is not intelligent. For some basic and routine content, intelligent customer service can quickly retrieve and match answers, which has indeed greatly improved efficiency. However, once faced with slightly more complicated situations, intelligent customer service will not be able to understand user needs, and will answer templates and routines. After several "dead loop" operations, it not only wastes time, but also damages the consumer's experience. For enterprises, intelligent customer service saves a lot of manpower for enterprises, but also consumes the intangible assets such as goodwill and user loyalty accumulated by the enterprise over the years; therefore, it is necessary to design an intelligent user service platform based on big data to improve user service satisfaction and user intelligent service quality. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent user service platform based on big data to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent user service platform based on big data, comprising a service data analysis and management module, a server-side semantic analysis module and an intelligent service interactive feedback module, characterized in that: the service data analysis and management module is used to analyze and manage service data of enterprise business products; the server-side semantic analysis module is used for the intelligent user server to analyze the text semantics in user consulting services; the intelligent service interactive feedback module is used to parse user consulting questions from the semantics of the questions and retrieve information to form the final intelligent service interactive feedback, and the service data analysis and management module, the server-side semantic analysis module and the intelligent service interactive feedback module are electrically connected.

[0005] According to the above technical solution, the method steps for running the platform include: Step 1: Integrate the data, processes, resources, organizations and functions related to business products in different life cycles of enterprise business services into a unified intelligent user service platform to analyze and manage service data; Step 2: Based on the shallow semantic analysis of the user consultation text, the keywords are marked according to the semantic similarity and vector model, and the similarity between them is calculated; Step 3: Construct rules for the intelligent user server to provide progressive services to users; Step 4: Analyze the semantics of user consultation questions and retrieve information to form the final intelligent service interaction feedback.

[0006] According to the above technical solution, the steps of analyzing and managing service data include: Through the intelligent service methods of intelligent perception connection analysis, decision optimization and execution control, an intelligent product service network is built, and the real demand item set of service users is analyzed and mined. In the intelligent services for business products, the basic data of service users in the entire life cycle business chain of enterprise business product services, including user basic feature data and business product preference data, is obtained through intelligent perception and service interaction data. The internal factor correlation relationship between the basic data is established, and the correlation analysis matrix is ​​constructed according to the correlation relationship. The potential service needs of service users are made explicit using the service user potential demand mining method based on fuzzy cognition. When the user enters the intelligent user service link, synchronous data information analysis is performed based on the correlation relationship. The service issues that the user may want to provide are predicted through the existing data. By providing predictive recommendation services in the form of related questions, a service list of questions that the user may ask is actively provided.

[0007] According to the above technical solution, the step of marking keywords based on the semantic similarity and vector model and calculating the similarity between them on the basis of the shallow semantic analysis of the user consultation text includes: The intelligent user service end obtains the overall transaction status and order information interaction data of the user group for the business product, pushes the service questions with high consultation frequency under the same historical status, obtains the service text entered by the user, performs word segmentation processing on the text entered by the user, and extracts the keywords in the text, specifically including: when the user makes a service consultation request to the intelligent user service end, the server end performs word segmentation and stop word removal preprocessing on the user's question sentence according to the professional vocabulary and business product content vocabulary, obtains the feature phrases corresponding to the question sentence, and then brings the obtained feature phrases into the classifier established by the server end to determine the category to which the user's query sentence belongs, and obtains the service candidate question set of the business product according to the combination of the synonym vocabulary and the keyword phrase; The consultation text input by the user is segmented, and the service knowledge graph is matched to the corresponding service database unit. The two service texts to be compared for similarity are preprocessed in the same way, including professional word tagging, segmentation, stop word removal, and part-of-speech tagging. All word units are randomly arranged to form a word sequence, and then the number of times the word units in the word sequence appear in the two service texts is counted respectively. The word unit is used as the vector dimension and the number of occurrences is the value of the corresponding dimension to obtain two vectors. Then the cosine value of the angle between the two vectors is calculated. The closer the cosine value is to 1, the more similar the two service texts are, and the closer it is to 0, the smaller the similarity between the two service texts is. By screening and identifying the keyword features of business product services, it is determined whether the service problem consulted by the user is related to non-business product services. When the judgment is no, the service is connected to the intelligent service window API of the non-business service, and the current business product intelligent service process is completed after obtaining the API feedback message. When the judgment is yes, the business product service problem consulted by the user is intelligently classified, and the service knowledge graph of the business product is used to match it to the corresponding service database unit. The information data of the consultation problem is retrieved and matched in the service knowledge graph of the business product, and the similarity with the service content of the constructed intelligent user service is calculated, and the service content is sorted and output based on the similarity.

[0008] According to the above technical solution, the step of constructing rules for the intelligent user server to provide progressive services to users includes: The dialogue process framework is constructed for different service dialogue contents of the intelligent customer service terminal of the business product, and the intelligent user service dialogue process is designed into a tree structure. According to the answers of the user and the intelligent customer service terminal, different subtrees are jumped. For open questions, answers are found from the existing knowledge base. If no suitable answers are found, it is jumped to the manual customer service. Before talking with the user, the basic information of the user and the business is first found from the database. The user name, telephone number, business information, business name, and package price information that were originally vacant in the opening remarks, prompts, and retention statements in the dialogue are added to the service answer, and then the dialogue begins. When the dialogue proceeds to the knowledge base question and answer, the service terminal counts the consulting questions of the service users and feeds back the statistical results to the semantic matching terminal. The semantic matching terminal will mark the priority of the questions in the knowledge base according to the feedback results. The higher the frequency of the question, the higher the priority. The service answers of the questions with high priority will be returned first in the next search and match. The feedback cycle of the intelligent user service effect is set to continuously iterate and optimize the service cycle during the service process. The tuning threshold is set according to the historical service performance of the intelligent user service. When the similarity reaches the highest point of the set threshold, the service answer to the user's consultation question is directly returned; when the similarity is at the lowest value of the set threshold, it is further determined whether the retrieved matching service content matches the user's consultation question, and whether the threshold point for automatically transferring to manual customer service to provide manual service to the user is determined. After the service is completed, the user's consultation question is located and the user consultation question dictionary is configured by improving the service knowledge graph of the business product. If the similarity is between the middle value of the threshold point of the set threshold, a list of candidate business product service consultation content is output to the user for independent selection. After the user makes a selection or the user re-enters the content according to the prompt of the candidate business product service consultation content list, the answer to the corresponding question is returned.

[0009] According to the above technical solution, the steps of analyzing the user's consultation questions from the perspective of question semantics and retrieving information include: When the user does not actively input keywords, but clicks on keywords or questions listed in advance by the intelligent customer service, the intelligent customer service terminal will display the corresponding service answers in the chat content, and the user will see hot keywords updated in real time on the interactive page. When the keywords or questions entered by the user do not have relevant answers in the service knowledge graph of the business product, the user will be automatically redirected to the manual customer service. After the redirection is completed, the intelligent customer service terminal will record the conversation between the manual customer service and the user, analyze the interactive service content, add the question-answer pairs that do not exist in the existing knowledge base to the existing knowledge base, and generate a knowledge base addition record. The intelligent customer service terminal management user can view the self-learning records of the intelligent customer service terminal through the records, correct the errors therein, and enable the intelligent customer service terminal to better adapt to business needs. After jumping to the manual customer service, in addition to recording the conversation between the manual customer service and the user, the system will continue to match the user's questions with the questions in the knowledge base. When there is an answer, the answer will be returned to the manual customer service for reference. When there is no answer, the user's question and the manual customer service's answer will be added to the knowledge base. An intervention threshold for active transfer to manual service is established. For users who have consulted in the intelligent link but ultimately failed to resolve the issue and were transferred to manual service, when the same question has been asked twice in a row, a service process is established to transfer the user to manual service when the user asks the third question. During the conversation between manual customer service and the user, when the user's consultation question has an answer in the knowledge base, the answer will be provided directly to the manual customer service. After access to manual service, rules for queuing will be set during peak periods or events, and queuing time will be set according to the number of users visiting in different time periods and the service users on the server. After waiting in line, the user will be connected to the corresponding manual customer service end according to the business scope pre-selected by the consumer user.

[0010] According to the above technical solution, the method steps for forming the final intelligent service interaction feedback include: the intelligent user service terminal simultaneously maintains a status table of manual customer service in the system, and the business field and current status of each manual customer service terminal are marked in the status table. When the service jumps from the intelligent service to the manual service, the status table is first checked to find the manual customer service related to the business and then the manual customer service that is in an idle state is screened out. The system jumps to the first manual customer service among the manual customer services that are finally screened out. When there are no idle manual customer service personnel on the manual customer service terminal of this business product, the user is prompted to wait and the user's inquiry problem data information and related business content are recorded.

[0011] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the present invention integrates data, processes, resources, organizations and functions related to business products in different life cycles of enterprise business services into a unified intelligent user service platform to analyze and manage service data, annotates keywords according to semantic similarity and vector models on the basis of shallow semantic analysis of user consultation text, calculates the similarity between them, and constructs rules for the intelligent user service end to provide hierarchical progressive services to users, and forms the final intelligent service interaction feedback from the semantic analysis of user consultation questions and information retrieval; solves the problem of invalid service caused by templated and routine answers of intelligent customer service in complex situations, improves user service experience and user service satisfaction, and also improves the quality of intelligent services of enterprises in intelligent customer service. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 A flowchart of a method for operating an intelligent user service platform based on big data provided in the first embodiment of the present invention; Figure 2 A schematic diagram of the module composition of an intelligent user service platform based on big data provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0013] 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.

[0014] Embodiment 1: Figure 1This is a flowchart of a method for operating a smart user service platform based on big data provided in the first embodiment of the present invention. This embodiment can be applied to the scenario of enterprise smart user services. The method can be executed by a smart user service platform based on big data provided in this embodiment, such as Figure 1 As shown, the method specifically comprises the following steps: Step 1: Integrate the data, processes, resources, organizations and functions related to business products in different life cycles of enterprise business services into a unified intelligent user service platform to analyze and manage service data; In the embodiment of the present invention, an intelligent product service network is constructed through an intelligent service method of intelligent perception connection analysis, decision optimization and execution control, and the real demand item set of service users is analyzed and mined. In the intelligent service for business products, the basic data of service users including basic feature data of users and business product preference data in the business chain of the entire life cycle of enterprise business product services are obtained through intelligent perception and service interaction data, and the internal factor association relationship between the basic data is established. According to the association relationship, an association analysis matrix is ​​constructed, and the potential service needs of service users are made explicit by using a service user potential demand mining method based on fuzzy cognition. When the user enters the intelligent user service link, synchronous data information analysis is performed based on the association relationship, and the service problems that the user may want to provide are predicted through existing data. By providing a prediction recommendation service in the form of associated problems, a service list of questions that the user may ask is actively provided; For example, the intelligent user service terminal sets up associated entries for maintenance of keywords with high relevance in service questions, and conducts regular analysis of intelligent user service topics to explore points for improving service relevance. When a user mentions one of the keywords, Lenovo predicts the consulting questions that the user will have regarding business products, and proactively provides service information push to the user. By introducing mature industry standards for intelligent customer service reliability and adding a mechanism for proactively monitoring the reliability of intelligent services, a reliability measurement standard for intelligent customer service in the intelligent user service center is constructed. After the standard is established, service perception is provided to users through decision tree analysis and using a variety of similarity algorithms between basic data. Step 2: Based on the shallow semantic analysis of the user consultation text, the keywords are marked according to the semantic similarity and vector model, and the similarity between them is calculated; In an embodiment of the present invention, the intelligent user service end obtains the overall transaction status and order information interaction data of the user group for the business product, pushes the service questions with high consultation frequency under the same historical status, obtains the service text input by the user, performs word segmentation processing on the text input by the user, and extracts keywords in the text, which specifically includes: when the user makes a service consultation request to the intelligent user service end, the server end performs word segmentation and stop word removal preprocessing on the user's question sentence according to the professional vocabulary and the business product content vocabulary to obtain the feature phrases corresponding to the question sentence, and then brings the obtained feature phrases into the classifier established by the server end to determine the category to which the user query sentence belongs, and obtains the service candidate question set of the business product according to the combination of the synonym vocabulary and the keyword phrase; Exemplarily, the consultation text input by the user is segmented, and the service knowledge graph is matched to the corresponding service database unit. The two service texts to be compared for similarity are preprocessed in the same way, including professional word tagging, segmentation, stop word removal, and part-of-speech tagging. All word units are randomly arranged to form a word sequence, and then the number of times the word units in the word sequence appear in the two service texts is counted respectively. The word unit is used as the vector dimension and the number of occurrences is used as the value of the corresponding dimension to obtain two vectors. The cosine value of the angle between the two vectors is then calculated. The closer the cosine value is to 1, the more similar the two service texts are, and the closer it is to 0, the smaller the similarity between the two service texts is. Exemplarily, by screening and identifying keyword features of business product services, it is determined whether the service problem consulted by the user is a problem related to non-business product services. When the judgment is no, the service is connected to the intelligent service window API of the non-business service, and the current business product intelligent service process is completed after obtaining the API feedback message. When the judgment is yes, the business product service problem consulted by the user is intelligently classified, and the service knowledge graph of the business product is used to match it to the corresponding service database unit. The information data of the consultation problem is retrieved and matched in the service knowledge graph of the business product, and the similarity with the service content of the constructed intelligent user service is calculated, and the service content is sorted and output according to the similarity.

[0015] Step 3: Construct rules for the intelligent user server to provide progressive services to users; In the embodiment of the present invention, a dialogue process framework is constructed for different service dialogue contents of the intelligent customer service terminal of the business product, and the intelligent user service dialogue process is designed into a tree structure. According to the answers of the user and the intelligent customer service terminal, the user jumps to different subtrees. For open questions, the answer is found from the existing knowledge base. If no suitable answer is found, the user jumps to the manual customer service. Before the dialogue with the user, the basic information of the user and the service is first found from the database, and the user name, telephone number, service information, service name, and package price information that are originally vacant in the opening remarks, prompts, and retention statements in the dialogue are added to the service answer, and then the dialogue begins. When the dialogue proceeds to the knowledge base question and answer, the service terminal counts the consulting questions of the service user and feeds back the statistical results to the semantic matching terminal. The semantic matching terminal will mark the priority of the questions in the knowledge base according to the feedback results. The higher the frequency of the question, the higher the priority. The service answer of the question with high priority will be returned first in the next search and matching. The feedback cycle of the intelligent user service effect is set, so that the service cycle is continuously iterated and optimized during the service process. Exemplarily, a tuning threshold is set according to the historical service performance of intelligent user services, and the service answer to the user's consultation question is directly returned when the similarity reaches the highest point of the set threshold; when the similarity is at the lowest value of the set threshold, it is further determined whether the retrieved matching service content matches the user's consultation question, and whether the threshold point for automatically transferring to manual customer service to provide manual service to the user is determined; after the service is completed, the user's consultation question is located and the user consultation question dictionary is configured by improving the service knowledge graph of the business product; if the similarity is between the middle value of the threshold point of the set threshold, a list of candidate business product service consultation content is output to the user for independent selection; after the user makes a selection or the user re-enters the content according to the prompt of the candidate business product service consultation content list, the answer to the corresponding question is returned.

[0016] Step 4: Analyze the semantics of user consultation questions and retrieve information to form the final intelligent service interaction feedback.

[0017] In an embodiment of the present invention, when the user does not actively input keywords, but clicks on keywords or questions listed in advance by the intelligent customer service, the intelligent customer service terminal displays the corresponding service answers in the chat content, and the user sees the hot keywords updated in real time on the interactive page. When the keywords or questions input by the user do not have relevant answers in the service knowledge graph of the business product, it automatically jumps to the manual customer service. After the jump is completed, the intelligent customer service terminal records the conversation between the manual customer service and the user, analyzes the interactive service content, adds the question-answer pairs that do not exist in the existing knowledge base to the existing knowledge base, and generates a knowledge base addition record. The intelligent customer service terminal management user views the self-learning records of the intelligent customer service terminal through the records, corrects the errors therein, and enables the intelligent customer service terminal to better adapt to business needs. After jumping to the manual customer service, in addition to recording the conversation between the manual customer service and the user, the system will continue to match the user's questions with the questions in the knowledge base. When there is an answer, the answer will be returned to the manual customer service for reference. When there is no answer, the user's question and the answer of the manual customer service are added to the knowledge base. Exemplarily, an intervention threshold for active transfer to a manual service end is set. For users who have consulted in the intelligent link but ultimately failed to resolve the issue and were transferred to manual service, when the same question has been asked twice in succession, when the user asks the question for the third time, a service process is established to transfer the user to manual service. During the conversation between the manual customer service and the user, when the user's consultation question has an answer in the knowledge base, it will be provided directly to the manual customer service. After accessing the manual service, rules for queuing are set during peak hours or activities, and the queuing time is set according to the number of users visiting in different time periods and the service users on the server end. After waiting in line, the user is connected to the corresponding manual customer service end according to the business scope pre-selected by the consumer user. The customer service staff communicates with the consumer through the system in the background, and improves the user's feedback experience on intelligent customer service and manual service by timely supplementing the manual service intervention; Exemplarily, the intelligent user service terminal also maintains a status table of manual customer service in the system, in which the business field and current status of each manual customer service terminal are marked. When the service jumps from the intelligent service to the manual service, the status table is first checked to find the manual customer service related to the business and then the manual customer service that is idle is screened out. The system jumps to the first manual customer service among the manual customer service that is finally screened out. When there are no idle manual customer service personnel on the manual customer service terminal of this business product, the user is prompted to wait and the user's question data information and related business content are recorded.

[0018] Embodiment 2: Embodiment 2 of the present invention provides an intelligent user service platform based on big data. Figure 2 A schematic diagram of the module composition of an intelligent user service platform based on big data provided in the second embodiment of the present invention, such as Figure 2As shown, the platform includes: Service data analysis and management module, used to analyze and manage service data of enterprise business products; The server-side semantic analysis module is used by the intelligent user server to analyze the text semantics in the user consultation service; The intelligent service interaction feedback module is used to parse user consultation questions from the semantics of the questions and retrieve information to form the final intelligent service interaction feedback.

[0019] In some embodiments of the present invention, the service data analysis management module includes: The service user demand item collection module is used to build an intelligent product service network through intelligent service methods such as intelligent perception connection analysis, decision optimization and execution control, and analyze and mine the real demand item collection of service users; Basic data acquisition module, used to acquire basic data of service users in the entire life cycle of enterprise business product service, including basic user feature data and business product preference data; The service demand explicitation module is used to make the potential service demands of service users explicit by using the service user potential demand mining method based on fuzzy cognition.

[0020] In some embodiments of the present invention, the server-side semantic analysis module includes: The keyword tagging module is used to tag keywords based on the semantic similarity and vector model based on the shallow semantic analysis of user consultation text; The feature phrase acquisition module is used for the server to segment the user's question and remove stop words according to the professional vocabulary and business product content vocabulary to obtain the feature phrase corresponding to the question; A word segmentation processing module is used to perform word segmentation processing on the consultation text input by the user; The keyword feature screening and identification module is used to screen and identify the keyword features of business products and services.

[0021] In some embodiments of the present invention, the intelligent service interaction feedback module includes: The progressive service rule building module is used to build rules for the intelligent user server to provide progressive services to the user level; The semantic matching terminal feedback module is used for the semantic matching terminal to mark the priority of the questions in the knowledge base according to the feedback results; A tuning threshold setting module, used to set a tuning threshold according to the historical service effect of the intelligent user service; The customer service terminal self-learning module is used for intelligent customer service terminals to record the conversations between human customer service and users, and analyze and self-learn the interactive service content.

[0022] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0023] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent user service platform based on big data, including a service data analysis and management module, a server-side semantic analysis module and an intelligent service interaction feedback module, characterized in that: The service data analysis and management module is used to analyze and manage the service data of the enterprise business products; the server-side semantic analysis module is used for the intelligent user server to analyze the text semantics in the user consulting service; the intelligent service interactive feedback module is used to parse the user consulting questions from the semantics of the questions and retrieve information to form the final intelligent service interactive feedback. The service data analysis and management module, the server-side semantic analysis module and the intelligent service interactive feedback module are electrically connected.

2. The intelligent user service platform based on big data according to claim 1, characterized in that: The service data analysis and management module includes: a service user demand item collection module, a basic data acquisition module, and a service demand explicitation module. The service user demand item collection module is used to build an intelligent product service network through an intelligent service method of intelligent perception connection analysis, decision optimization and execution control, and analyze and mine the real demand item collection of service users; the basic data acquisition module is used to obtain basic data of service users in the entire life cycle business chain of enterprise business products and services, including user basic feature data and business product preference data; the service demand explicitation module is used to make the potential service needs of service users explicit by using a service user potential demand mining method based on fuzzy cognition. The service user demand item collection module, the basic data acquisition module and the service demand explicitation module are electrically connected.

3. The intelligent user service platform based on big data according to claim 2, characterized in that: The server-side semantic analysis module includes: a keyword tagging module, a feature phrase acquisition module, a word segmentation processing module and a keyword feature screening and identification module. The keyword tagging module is used to tag keywords according to semantic similarity and vector model based on shallow semantic analysis of user consultation text; the feature phrase acquisition module is used for the server to segment and remove stop words from user questions according to professional vocabulary and business product content vocabulary to obtain feature phrases corresponding to the questions; the word segmentation processing module is used to segment the consultation text input by the user; the keyword feature screening and identification module is used to screen and identify keyword features of business product services, the keyword tagging module is electrically connected to the feature phrase acquisition module, and the word segmentation processing module is electrically connected to the keyword feature screening and identification module.

4. The intelligent user service platform based on big data according to claim 3, characterized in that: The intelligent service interactive feedback module includes: a progressive service rule construction module, a semantic matching terminal feedback module, a tuning threshold setting module and a customer service terminal self-learning module. The progressive service rule construction module is used to construct rules for the intelligent user service end to provide progressive services to the user level; the semantic matching terminal feedback module is used for the semantic matching terminal to prioritize the questions in the knowledge base according to the feedback results; the tuning threshold setting module is used to set the tuning threshold according to the historical service effect of the intelligent user service; the customer service terminal self-learning module is used for the intelligent customer service terminal to record the dialogue between the manual customer service and the user, analyze and self-learn the interactive service content, the progressive service rule construction module is electrically connected to the semantic matching terminal feedback module, and the tuning threshold setting module is electrically connected to the customer service terminal self-learning module.

5. The intelligent user service platform based on big data according to claim 4, characterized in that: The method steps for running the platform include: Step 1: Integrate the data, processes, resources, organizations and functions related to business products in different life cycles of enterprise business services into a unified intelligent user service platform to analyze and manage service data; Step 2: Based on the shallow semantic analysis of the user consultation text, the keywords are marked according to the semantic similarity and vector model, and the similarity between them is calculated; Step 3: Construct rules for the intelligent user server to provide progressive services to users; Step 4: Analyze the semantics of user consultation questions and retrieve information to form the final intelligent service interaction feedback.

6. The intelligent user service platform based on big data according to claim 5, characterized in that: The steps of analyzing and managing the service data include: Through the intelligent service methods of intelligent perception connection analysis, decision optimization and execution control, an intelligent product service network is built, and the real demand item set of service users is analyzed and mined. In the intelligent services for business products, the basic data of service users in the entire life cycle business chain of enterprise business product services, including user basic feature data and business product preference data, is obtained through intelligent perception and service interaction data. The internal factor correlation relationship between the basic data is established, and the correlation analysis matrix is ​​constructed according to the correlation relationship. The potential service needs of service users are made explicit using the service user potential demand mining method based on fuzzy cognition. When the user enters the intelligent user service link, synchronous data information analysis is performed based on the correlation relationship. The service issues that the user may want to provide are predicted through the existing data. By providing predictive recommendation services in the form of related questions, a service list of questions that the user may ask is actively provided.

7. The intelligent user service platform based on big data according to claim 6, characterized in that: The step of marking keywords based on the semantic similarity and vector model and calculating the similarity between them on the basis of the shallow semantic analysis of the user consultation text includes: The intelligent user service end obtains the overall transaction status and order information interaction data of the user group for the business product, pushes the service questions with high consultation frequency under the same historical status, obtains the service text entered by the user, performs word segmentation processing on the text entered by the user, and extracts the keywords in the text, specifically including: when the user makes a service consultation request to the intelligent user service end, the server end performs word segmentation and stop word removal preprocessing on the user's question sentence according to the professional vocabulary and business product content vocabulary, obtains the feature phrases corresponding to the question sentence, and then brings the obtained feature phrases into the classifier established by the server end to determine the category to which the user's query sentence belongs, and obtains the service candidate question set of the business product according to the combination of the synonym vocabulary and the keyword phrase; The consultation text input by the user is segmented, and the service knowledge graph is matched to the corresponding service database unit. The two service texts to be compared for similarity are preprocessed in the same way, including professional word tagging, segmentation, stop word removal, and part-of-speech tagging. All word units are randomly arranged to form a word sequence, and then the number of times the word units in the word sequence appear in the two service texts is counted respectively. The word unit is used as the vector dimension and the number of occurrences is the value of the corresponding dimension to obtain two vectors. Then the cosine value of the angle between the two vectors is calculated. The closer the cosine value is to 1, the more similar the two service texts are, and the closer it is to 0, the smaller the similarity between the two service texts is. By screening and identifying the keyword features of business product services, it is determined whether the service problem consulted by the user is related to non-business product services. When the judgment is no, the service is connected to the intelligent service window API of the non-business service, and the current business product intelligent service process is completed after obtaining the API feedback message. When the judgment is yes, the business product service problem consulted by the user is intelligently classified, and the service knowledge graph of the business product is used to match it to the corresponding service database unit. The information data of the consultation problem is retrieved and matched in the service knowledge graph of the business product, and the similarity with the service content of the constructed intelligent user service is calculated, and the service content is sorted and output based on the similarity.

8. The intelligent user service platform based on big data according to claim 7, characterized in that: The step of constructing rules for the intelligent user server to provide progressive services to users includes: The dialogue process framework is constructed for different service dialogue contents of the intelligent customer service terminal of the business product, and the intelligent user service dialogue process is designed into a tree structure. According to the answers of the user and the intelligent customer service terminal, different subtrees are jumped. For open questions, answers are found from the existing knowledge base. If no suitable answers are found, it is jumped to the manual customer service. Before talking with the user, the basic information of the user and the business is first found from the database. The user name, telephone number, business information, business name, and package price information that were originally vacant in the opening remarks, prompts, and retention statements in the dialogue are added to the service answer, and then the dialogue begins. When the dialogue proceeds to the knowledge base question and answer, the service terminal counts the consulting questions of the service users and feeds back the statistical results to the semantic matching terminal. The semantic matching terminal will mark the priority of the questions in the knowledge base according to the feedback results. The higher the frequency of the question, the higher the priority. The service answers of the questions with high priority will be returned first in the next search and match. The feedback cycle of the intelligent user service effect is set to continuously iterate and optimize the service cycle during the service process. The tuning threshold is set according to the historical service performance of the intelligent user service. When the similarity reaches the highest point of the set threshold, the service answer to the user's consultation question is directly returned; when the similarity is at the lowest value of the set threshold, it is further determined whether the retrieved matching service content matches the user's consultation question, and whether the threshold point for automatically transferring to manual customer service to provide manual service to the user is determined. After the service is completed, the user's consultation question is located and the user consultation question dictionary is configured by improving the service knowledge graph of the business product. If the similarity is between the middle value of the threshold point of the set threshold, a list of candidate business product service consultation content is output to the user for independent selection. After the user makes a selection or the user re-enters the content according to the prompt of the candidate business product service consultation content list, the answer to the corresponding question is returned.

9. The intelligent user service platform based on big data according to claim 8, characterized in that: The steps of analyzing the user's inquiry questions from the perspective of question semantics and retrieving information include: When the user does not actively input keywords, but clicks on keywords or questions listed in advance by the intelligent customer service, the intelligent customer service terminal will display the corresponding service answers in the chat content, and the user will see hot keywords updated in real time on the interactive page. When the keywords or questions entered by the user do not have relevant answers in the service knowledge graph of the business product, the user will be automatically redirected to the manual customer service. After the redirection is completed, the intelligent customer service terminal will record the conversation between the manual customer service and the user, analyze the interactive service content, add the question-answer pairs that do not exist in the existing knowledge base to the existing knowledge base, and generate a knowledge base addition record. The intelligent customer service terminal management user can view the self-learning records of the intelligent customer service terminal through the records, correct the errors therein, and enable the intelligent customer service terminal to better adapt to business needs. After jumping to the manual customer service, in addition to recording the conversation between the manual customer service and the user, the system will continue to match the user's questions with the questions in the knowledge base. When there is an answer, the answer will be returned to the manual customer service for reference. When there is no answer, the user's question and the manual customer service's answer will be added to the knowledge base. An intervention threshold for active transfer to manual service is established. For users who have consulted in the intelligent link but ultimately failed to resolve the issue and were transferred to manual service, when the same question has been asked twice in a row, a service process is established to transfer the user to manual service when the user asks the third question. During the conversation between manual customer service and the user, when the user's consultation question has an answer in the knowledge base, the answer will be provided directly to the manual customer service. After access to manual service, rules for queuing will be set during peak periods or events, and queuing time will be set according to the number of users visiting in different time periods and the service users on the server. After waiting in line, the user will be connected to the corresponding manual customer service end according to the business scope pre-selected by the consumer user.

10. The intelligent user service platform based on big data according to claim 9, characterized in that: The method steps for forming the final intelligent service interactive feedback include: the intelligent user service terminal simultaneously maintains a status table of manual customer service in the system, in which the business field and current status of each manual customer service terminal are marked; when the service jumps from the intelligent service to the manual service, the status table is first checked to find the manual customer service related to the business, and then the manual customer service in the idle state is screened out; the system jumps to the first manual customer service among the manual customer services screened out at last; when there are no idle manual customer service personnel at the manual customer service terminal of this business product, the user is prompted to wait and the user's inquiry data information and related business content are recorded.

Images