B2B system intelligent customer service management and control method and system based on natural language processing

Abstract

The invention discloses a B2B system intelligent customer service management and control method and system based on natural language processing. The method comprises the steps that natural language query information fed back by a user through multiple channels and based on a B2B platform is acquired; performing preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used for converting the natural language query information into text data which can be recognized; executing target processing operation on the preprocessing result to obtain a target processing result aiming at the preprocessing result; the target processing operation comprises at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model and a third information analysis operation based on a knowledge base; the target processing result is used for indicating a response item responding to the natural language query information. Therefore, the response intelligence degree, accuracy and humanization degree of the query information can be improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent customer service technology, and in particular to a B2B system intelligent customer service management and control method and system based on natural language processing. Background Art

[0002] In the existing B2B intelligent customer service field, with the increasing frequency of business exchanges between enterprises, users are increasingly dependent on intelligent customer service. However, current intelligent customer service systems face many difficulties when handling user inquiries.

[0003] On the one hand, there are numerous channels for user feedback, such as email, online customer service windows, and phone messages. However, the information from each channel is independent and lacks an effective integration mechanism. This forces customer service personnel to spend a lot of time and energy collecting and organizing a large amount of scattered information. This is not only inefficient, but also prone to information omissions, making it difficult to fully understand user issues, which in turn affects service quality.

[0004] On the other hand, natural language query information is highly complex and diverse. Users' expression habits and language styles vary, and queries may involve specialized domain knowledge and complex business logic. Existing intelligent customer service systems often lack effective pre-processing methods when processing this information, making it difficult to accurately convert natural language into a data format that computers can understand and process. This leads to deviations in subsequent analysis and response processes, making it difficult to accurately capture user intent.

[0005] In addition, when responding to user inquiries, conventional intelligent customer service cannot give accurate and comprehensive responses to some complex or ambiguous questions, and cannot meet the diverse needs of users, which reduces users' satisfaction and trust in intelligent customer service. Summary of the Invention

[0006] The present invention provides a B2B system intelligent customer service management and control method and system based on natural language processing, which can realize the accurate collection of multi-channel query information and improve the intelligence, accuracy and humanization of the response to the query information.

[0007] In order to solve the above technical problems, the first aspect of the present invention discloses a B2B system intelligent customer service management and control method based on natural language processing, the method comprising:

[0008] Obtaining natural language query information fed back by users through multiple channels, wherein the natural language query information includes query information triggered by the users through the B2B platform;

[0009] Performing a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data;

[0010] A target processing operation is performed on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to the natural language query information.

[0011] As an optional implementation manner, in the first aspect of the present invention, performing a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result includes:

[0012] performing a first information analysis operation on the preprocessing result according to the context understanding module to obtain a first information analysis result corresponding to the preprocessing result; the first information analysis operation includes at least one of a history query, a session state analysis, and a user preference analysis;

[0013] performing a second information analysis operation on the first information analysis result according to the semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result; the second information analysis operation includes at least one of user query intent recognition, entity information parsing, and user sentiment tendency analysis;

[0014] According to the knowledge base, a third information analysis operation is performed on the second information analysis result to obtain a third information analysis result corresponding to the second information analysis result as the target processing result; the third information analysis operation is used to match the response content in response to the second information analysis result.

[0015] As an optional implementation, in the first aspect of the present invention, the context understanding module performs a first information analysis operation on the preprocessing result to obtain a first information analysis result corresponding to the preprocessing result, including:

[0016] According to a preset session state model, real-time monitoring of the session state after the user triggers the natural language query information, wherein the session state includes any one of the states of session start, ongoing, paused, and ended;

[0017] When it is determined that the session state is the session start or the session in progress, detecting whether there is a historical conversation record corresponding to the preprocessing result; and when the historical conversation record is detected to exist, performing information extraction on the historical conversation record according to the context understanding module to obtain a record summary corresponding to the historical conversation record, the record summary including at least one of the historical conversation content, the question type, and the solution, the correlation degree of which to the natural language query information is higher than a preset correlation threshold;

[0018] Performing user preference analysis on the historical conversation records according to the context understanding module to obtain the user's query preference items, potential demand items, and points of interest items as user preference information;

[0019] The record summary and the user preference information are associated and integrated with the preprocessing result to update the preprocessing result and obtain a first information analysis result.

[0020] As an optional embodiment, in the first aspect of the present invention, performing a second information analysis operation on the first information analysis result according to the semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result includes:

[0021] Performing user query intent recognition on the first information analysis result according to the semantic analysis model to obtain the user's intent recognition result; the intent recognition result includes an intent category and a confidence score;

[0022] performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result; the entity information parsing is used to extract keywords corresponding to the information query in the preprocessing result; the target entity information includes one or more entity keywords;

[0023] Performing sentiment analysis on the first information analysis result according to the semantic analysis model to obtain the target sentiment of the user triggering the natural language query information and its corresponding sentiment intensity; the target sentiment may include positive sentiment, negative sentiment, or neutral sentiment;

[0024] The intention recognition result, the target entity information, the target emotion and its corresponding emotion intensity are determined as a second information analysis result corresponding to the first information analysis result.

[0025] As an optional implementation manner, in the first aspect of the present invention, performing a third information analysis operation on the second information analysis result according to the knowledge base to obtain a third information analysis result corresponding to the second information analysis result includes:

[0026] performing a first-level matching operation on the second information analysis result according to the knowledge base to obtain a first-level matching result corresponding to the second information analysis result; the first-level matching operation includes knowledge item retrieval; the first-level matching result includes at least one knowledge item to be screened;

[0027] performing a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result; the secondary matching operation includes in-depth analysis, matching degree calculation, applicability scoring and ranking based on the first information analysis result and the second information analysis result; the secondary matching result includes target knowledge items that meet the screening requirements;

[0028] According to the second information analysis result, information filling and emotion adjustment are performed on the secondary matching result to obtain a target output result as the third information analysis result.

[0029] As an optional implementation, in the first aspect of the present invention, performing a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result includes:

[0030] For each of the knowledge items to be screened, performing a text similarity calculation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a similarity calculation result between the second information analysis result and the knowledge item to be screened; the text similarity calculation operation includes a text preprocessing operation, a feature extraction operation, and a similarity calculation operation;

[0031] Performing a semantic annotation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a semantic annotation result corresponding to the second information analysis result and the knowledge item to be screened; the semantic annotation operation includes a syntactic structure parsing operation, a semantic role annotation operation, and a semantic role comparison operation;

[0032] Based on the similarity calculation result and the semantic annotation result corresponding to the knowledge item to be screened, a comprehensive similarity score between the second information analysis result and the knowledge item to be screened is calculated, and a suitability score corresponding to the knowledge item to be screened and the user's query requirements and / or conversation scenario is calculated;

[0033] Based on the comprehensive similarity score and the applicability score corresponding to each of the knowledge items to be screened, all the knowledge items to be screened are scored and sorted, and any of the knowledge items to be screened that is ranked within the preset ranking range is selected from the corresponding scoring and sorting results, recorded as the target knowledge item, and used as the secondary matching result.

[0034] As an optional embodiment, in the first aspect of the present invention, the text preprocessing operation includes word segmentation, stop word removal, stemming, and phrase reorganization; the text preprocessing is used to process the second information analysis result into a standardized text representation;

[0035] The semantic role labeling operation is used to label all semantic roles corresponding to the second information analysis result, and all the semantic roles include agent role, patient role, tool, time and place.

[0036] A second aspect of the present invention discloses a B2B system intelligent customer service management and control system based on natural language processing, the system comprising:

[0037] An acquisition module, configured to acquire natural language query information fed back by users through multiple channels, wherein the natural language query information includes query information triggered by the users through the B2B platform;

[0038] A preprocessing module, configured to perform a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is configured to convert the natural language query information into recognizable text data;

[0039] A target processing module is used to perform a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to the natural language query information.

[0040] As an optional implementation manner, in the second aspect of the present invention, the target processing module includes:

[0041] a first information analysis submodule configured to perform a first information analysis operation on the preprocessing result according to the context understanding module to obtain a first information analysis result corresponding to the preprocessing result; the first information analysis operation comprising at least one of a history record query, a session state analysis, and a user preference analysis;

[0042] A second information analysis submodule is configured to perform a second information analysis operation on the first information analysis result according to a semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result; the second information analysis operation includes at least one of user query intent recognition, entity information parsing, and user sentiment tendency analysis;

[0043] The third information analysis submodule is used to perform a third information analysis operation on the second information analysis result according to the knowledge base, and obtain a third information analysis result corresponding to the second information analysis result as the target processing result; the third information analysis operation is used to match the response content in response to the second information analysis result.

[0044] As an optional embodiment, in the second aspect of the present invention, the first information analysis submodule performs a first information analysis operation on the preprocessing result according to the context understanding module, and a method for obtaining a first information analysis result corresponding to the preprocessing result specifically includes:

[0045] According to a preset session state model, real-time monitoring of the session state after the user triggers the natural language query information, wherein the session state includes any one of the states of session start, ongoing, paused, and ended;

[0046] When it is determined that the session state is the session start or the session in progress, detecting whether there is a historical conversation record corresponding to the preprocessing result; and when the historical conversation record is detected to exist, performing information extraction on the historical conversation record according to the context understanding module to obtain a record summary corresponding to the historical conversation record, the record summary including at least one of the historical conversation content, the question type, and the solution, the correlation degree of which to the natural language query information is higher than a preset correlation threshold;

[0047] Performing user preference analysis on the historical conversation records according to the context understanding module to obtain the user's query preference items, potential demand items, and points of interest items as user preference information;

[0048] The record summary and the user preference information are associated and integrated with the preprocessing result to update the preprocessing result and obtain a first information analysis result.

[0049] As an optional embodiment, in the second aspect of the present invention, the second information analysis submodule performs a second information analysis operation on the first information analysis result according to the semantic analysis model, and a method for obtaining a second information analysis result corresponding to the first information analysis result specifically includes:

[0050] Performing user query intent recognition on the first information analysis result according to the semantic analysis model to obtain the user's intent recognition result; the intent recognition result includes an intent category and a confidence score;

[0051] performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result; the entity information parsing is used to extract keywords corresponding to the information query in the preprocessing result; the target entity information includes one or more entity keywords;

[0052] Performing sentiment analysis on the first information analysis result according to the semantic analysis model to obtain the target sentiment of the user triggering the natural language query information and its corresponding sentiment intensity; the target sentiment may include positive sentiment, negative sentiment, or neutral sentiment;

[0053] The intention recognition result, the target entity information, the target emotion and its corresponding emotion intensity are determined as a second information analysis result corresponding to the first information analysis result.

[0054] As an optional embodiment, in the second aspect of the present invention, the third information analysis submodule performs a third information analysis operation on the second information analysis result according to the knowledge base, and a method for obtaining a third information analysis result corresponding to the second information analysis result specifically includes:

[0055] performing a first-level matching operation on the second information analysis result according to the knowledge base to obtain a first-level matching result corresponding to the second information analysis result; the first-level matching operation includes knowledge item retrieval; the first-level matching result includes at least one knowledge item to be screened;

[0056] performing a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result; the secondary matching operation includes in-depth analysis, matching degree calculation, applicability scoring and ranking based on the first information analysis result and the second information analysis result; the secondary matching result includes target knowledge items that meet the screening requirements;

[0057] According to the second information analysis result, information filling and emotion adjustment are performed on the secondary matching result to obtain a target output result as the third information analysis result.

[0058] As an optional embodiment, in the second aspect of the present invention, the third information analysis submodule performs a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result, specifically comprising:

[0059] For each of the knowledge items to be screened, performing a text similarity calculation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a similarity calculation result between the second information analysis result and the knowledge item to be screened; the text similarity calculation operation includes a text preprocessing operation, a feature extraction operation, and a similarity calculation operation;

[0060] Performing a semantic annotation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a semantic annotation result corresponding to the second information analysis result and the knowledge item to be screened; the semantic annotation operation includes a syntactic structure parsing operation, a semantic role annotation operation, and a semantic role comparison operation;

[0061] Based on the similarity calculation result and the semantic annotation result corresponding to the knowledge item to be screened, a comprehensive similarity score between the second information analysis result and the knowledge item to be screened is calculated, and a suitability score corresponding to the knowledge item to be screened and the user's query requirements and / or conversation scenario is calculated;

[0062] Based on the comprehensive similarity score and the applicability score corresponding to each of the knowledge items to be screened, all the knowledge items to be screened are scored and sorted, and any of the knowledge items to be screened that is ranked within the preset ranking range is selected from the corresponding scoring and sorting results, recorded as the target knowledge item, and used as the secondary matching result.

[0063] As an optional embodiment, in the second aspect of the present invention, the text preprocessing operation includes word segmentation, stop word removal, stemming, and phrase reorganization; the text preprocessing is used to process the second information analysis result into a standardized text representation;

[0064] The semantic role labeling operation is used to label all semantic roles corresponding to the second information analysis result, and all the semantic roles include agent role, patient role, tool, time and place.

[0065] A third aspect of the present invention discloses a B2B system intelligent customer service management and control device based on natural language processing, the device comprising:

[0066] a memory storing executable program code;

[0067] a processor coupled to the memory;

[0068] The processor calls the executable program code stored in the memory to execute the B2B system intelligent customer service management and control method based on natural language processing disclosed in the first aspect of the present invention.

[0069] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the B2B system intelligent customer service management and control method based on natural language processing disclosed in the first aspect of the present invention.

[0070] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0071] In an embodiment of the present invention, a method for intelligent customer service management and control of a B2B system based on natural language processing is provided, the method comprising: obtaining natural language query information fed back by users through multiple channels, the natural language query information including query information triggered by users through a B2B platform; performing a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data; performing a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response matter in response to the natural language query information. It can be seen that the implementation of the present invention, by performing preprocessing on natural language query information and converting it into standardized and structured text data, can effectively remove redundant data and improve the processing efficiency and accuracy of the natural language query information; and, by combining the context understanding module, semantic analysis model and knowledge base, multi-dimensional and in-depth information analysis is performed on the preprocessing results, so as to accurately capture the user's query intention and explore the user's potential needs, which is conducive to improving the response accuracy and reliability of subsequent natural language query information triggered by the user and improving the user's query experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0073] Figure 1 This is a flow chart of a method for intelligent customer service management and control of a B2B system based on natural language processing disclosed in an embodiment of the present invention;

[0074] Figure 2 This is a flow chart of another method for intelligent customer service management and control of a B2B system based on natural language processing disclosed in an embodiment of the present invention;

[0075] Figure 3This is a schematic diagram of the structure of a B2B system intelligent customer service management and control system based on natural language processing disclosed in an embodiment of the present invention;

[0076] Figure 4 This is a schematic diagram of the structure of another B2B system intelligent customer service management and control system based on natural language processing disclosed in an embodiment of the present invention;

[0077] Figure 5 This is a structural diagram of a B2B system intelligent customer service management and control device based on natural language processing disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0078] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0079] The terms "first," "second," and so on, in the description and claims of the present invention and the accompanying drawings are used to distinguish between different items, not to describe a specific order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or end comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed therein, or may optionally include other steps or elements inherent to such process, method, product, or end.

[0080] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0081] The present invention discloses a B2B system intelligent customer service management and control method and system based on natural language processing. By performing preprocessing on natural language query information and converting it into standardized and structured text data, it can effectively remove redundant data and improve the processing efficiency and accuracy of the natural language query information; and by combining a context understanding module, a semantic analysis model and a knowledge base, a multi-dimensional and in-depth information analysis is performed on the preprocessing results, thereby accurately capturing the user's query intention and exploring the user's potential needs, which is conducive to improving the accuracy and reliability of subsequent responses to natural language query information triggered by the user and improving the user's query experience. The following are detailed descriptions.

[0082] Example 1

[0083] See also Figure 1 , Figure 1 This is a flow chart of a B2B system intelligent customer service management method based on natural language processing disclosed in an embodiment of the present invention. Figure 1 The described B2B system intelligent customer service management method based on natural language processing can be applied to a B2B system, which can further be a B2B system intelligent customer service management system based on natural language processing, and the embodiments of the present invention do not limit this. Figure 1 As shown, the B2B system intelligent customer service management and control method based on natural language processing may include the following operations:

[0084] 101. Obtain natural language query information fed back by users through multiple channels. The natural language query information includes query information triggered by users through the B2B platform.

[0085] In an embodiment of the present invention, the multiple channels may include online and offline channels. For online channels, natural language query information can be triggered / submitted through the online customer service window, message board function, and submission form on the B2B platform; or through official social media accounts, industry forums and communities, or email. For offline channels, natural language query information can be triggered / submitted through customer service phone calls, questionnaires filled out at offline industry exhibitions or seminars, etc., and the B2B platform can collect inquiry records from the customer service phone calls and read the uploaded questionnaire data.

[0086] In the embodiment of the present invention, the natural language query information refers to non-machine language triggered by the user, for example, "I want to inquire about the ab function on your platform" or "what is the cd function on your platform".

[0087] In the embodiment of the present invention, natural language query information triggered by users is obtained through multiple channels, so that various query information triggered by users through the B2B platform can be covered, which is conducive to improving the completeness and timeliness of the acquisition of the natural language query information.

[0088] 102. Perform a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data.

[0089] In an embodiment of the present invention, the preprocessing operation includes text vectorization, and also includes at least one operation of text cleaning, part-of-speech processing, part-of-speech tagging, and spelling correction.

[0090] In an embodiment of the present invention, the text vectorization is used to convert natural language query information into vector information that can be processed / recognized by a computer; and the text vectorization can be performed using a conventional bag of words model or a TF-IDF (Term Frequency-Inverse Document Frequency) model.

[0091] In embodiments of the present invention, text cleaning is used to filter out noise from natural language query information, such as HTML tags (if the information comes from a web form), special characters (such as emoticons and garbled characters), extra spaces, and line breaks. Text cleaning is also used to unify text formats, such as unifying full-width and half-width characters and uppercase and lowercase English.

[0092] In an embodiment of the present invention, the word segmentation process includes dictionary-based word segmentation and / or word segmentation combined with statistical methods; the dictionary-based word segmentation uses a pre-built word segmentation dictionary to segment the text according to certain rules. For example, for the Chinese text "I want to query order logistics information", the word segmentation dictionary contains words such as "I", "want", "query", "order", "logistics", and "information". The text can be segmented into "I / want / query / order / logistics / information" through dictionary matching. The word segmentation combined with statistical methods can be a statistical word segmentation based on the n-gram model. Specifically, by analyzing the frequency of occurrence of adjacent characters or words in the text, it is determined whether word segmentation should be performed. For example, for "Nanjing Yangtze River Bridge", it is found through statistics that the frequency of occurrence of "Nanjing City" and "Yangtze River Bridge" is higher, while the frequency of occurrence of combinations such as "Nanjing Mayor" and "River Bridge" is lower, so the text can be correctly segmented into "Nanjing City / Yangtze River Bridge".

[0093] In an embodiment of the present invention, the part-of-speech tagging can further tag the part-of-speech processing results based on the above part-of-speech processing, including tagging of nouns, verbs, and adjectives. For example, for the segmented sentences "I / want / query / order / logistics / information", "I" is tagged as a personal pronoun, "want" is tagged as a verb, "query" is tagged as a verb, "order" is tagged as a noun, "logistics" is tagged as a noun, and "information" is tagged as a noun.

[0094] In an embodiment of the present invention, the spelling correction can use an algorithm based on edit distance to calculate the edit distance between the word entered by the user and each word in the dictionary, and select the word with the smallest edit distance as the corrected result. For example, if the user enters "dingdan" (which may be a misspelling of "order" in the Chinese dictionary), by comparing it with the words in the dictionary, it is found that "order" has the smallest edit distance with "dingdan", and thus "dingdan" is corrected to "order".

[0095] 103. Perform a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to natural language query information.

[0096] It can be seen that implementation Figure 1 The described B2B system intelligent customer service management method based on natural language processing performs preprocessing on natural language query information and converts it into standardized and structured text data, which can effectively remove redundant data and improve the processing efficiency and accuracy of the natural language query information; and, by combining the context understanding module, semantic analysis model and knowledge base, conducts multi-dimensional and in-depth information analysis on the preprocessing results, so as to accurately capture the user's query intention and explore the user's potential needs, which is conducive to improving the accuracy and reliability of subsequent responses to natural language query information triggered by the user and improving the user's query experience.

[0097] Example 2

[0098] See also Figure 2 , Figure 2 This is a flow chart of another B2B system intelligent customer service management method based on natural language processing disclosed in an embodiment of the present invention. Figure 2 The described B2B system intelligent customer service management method based on natural language processing can be applied to a B2B system, which can further be a B2B system intelligent customer service management system based on natural language processing, and the embodiments of the present invention do not limit this. Figure 2As shown, the B2B system intelligent customer service management and control method based on natural language processing may include the following operations:

[0099] 201. Obtain natural language query information fed back by users through multiple channels, where the natural language query information includes query information triggered by users through the B2B platform.

[0100] 202. Perform a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data.

[0101] In the embodiment of the present invention, for other descriptions of step 201-step 202, please refer to other specific descriptions of step 101-step 102 in embodiment 1, and the embodiment of the present invention will not be repeated here.

[0102] 203. According to the context understanding module, perform a first information analysis operation on the preprocessing result to obtain a first information analysis result corresponding to the preprocessing result; the first information analysis operation includes at least one operation of history record query, session state analysis, and user preference analysis.

[0103] 204. According to the semantic analysis model, a second information analysis operation is performed on the first information analysis result to obtain a second information analysis result corresponding to the first information analysis result; the second information analysis operation includes at least one operation of user query intention recognition, entity information parsing and user sentiment tendency analysis.

[0104] 205. According to the knowledge base, perform a third information analysis operation on the second information analysis result to obtain a third information analysis result corresponding to the second information analysis result as the target processing result; the third information analysis operation is used to match the response content in response to the second information analysis result.

[0105] In the embodiment of the present invention, the response content may be based on a preset response template, and the third information analysis result may be filled in the response template and displayed.

[0106] It can be seen that implementation Figure 2The described method for intelligent customer service management and control of B2B systems based on natural language processing achieves efficient and accurate responses to user queries in the B2B system through a multi-level information analysis and processing process. Specifically, it can utilize a context understanding module, combined with historical record query, session state analysis, and user preference analysis, to comprehensively capture the contextual information of user queries, thereby improving the correlation between information analysis and context and, consequently, the accuracy of information analysis. Furthermore, it can utilize a semantic analysis model to deeply identify and analyze user query intent, entity information, and sentiment, thereby improving the accuracy of analyzing users' actual query needs. Furthermore, it can intelligently match the analysis results based on a knowledge base, quickly locating and generating responses that are highly consistent with user needs. This achieves efficient utilization of knowledge resources while increasing the intelligence of the response process corresponding to user query needs and improving service efficiency and quality.

[0107] In an optional embodiment, the context understanding module performs the first information analysis operation on the preprocessing result to obtain the first information analysis result corresponding to the preprocessing result, specifically including:

[0108] According to the preset session state model, real-time monitoring of the session state after the user triggers the natural language query information, the session state includes any of the states of session start, ongoing, paused, and ended;

[0109] When the session state is determined to be a session start or in progress, detecting whether there is a historical conversation record corresponding to the preprocessing result; when the historical conversation record is detected to exist, performing information extraction on the historical conversation record according to the context understanding module to obtain a record summary corresponding to the historical conversation record, the record summary including at least one of the historical conversation content, the question type, and the solution, the correlation degree of which to the natural language query information is higher than a preset correlation threshold;

[0110] The context understanding module performs user preference analysis on historical conversation records to obtain the user's query preference items, potential demand items, and points of interest items as user preference information;

[0111] The record summary and user preference information are associated and integrated with the preprocessing result to update the preprocessing result and obtain a first information analysis result.

[0112] It can be seen that in this optional embodiment, based on the preset session state model, real-time monitoring of the entire cycle of the user session can be achieved, and key nodes such as the start, progress, pause, and end of the session can be accurately identified, providing a dynamic context benchmark for subsequent information analysis, ensuring that the analysis process is highly synchronized with the session process; and, in the active phase of the session (start / progress), it can automatically retrieve and associate historical conversation records, and extract highly correlated summary information (including historical conversation content, problem type, solution, etc.) through the context understanding module, thereby effectively reusing historical interaction experience and improving the efficiency of current queries; and, it can also perform multi-dimensional preference analysis based on historical conversation records, accurately construct user query preferences, potential needs and interest point portraits, realize in-depth enhancement of personalized service capabilities, and further improve the matching degree and accuracy of response content with the user's real needs.

[0113] In another optional embodiment, the method of performing the second information analysis operation on the first information analysis result according to the semantic analysis model to obtain the second information analysis result corresponding to the first information analysis result specifically includes:

[0114] Performing user query intent recognition on the first information analysis result according to the semantic analysis model to obtain a user intent recognition result; the intent recognition result includes an intent category and a confidence score;

[0115] Performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result; the entity information parsing is used to extract keywords corresponding to the information query in the preprocessing result; the target entity information includes one or more entity keywords;

[0116] Performing sentiment analysis on the first information analysis result according to the semantic analysis model to obtain the target sentiment of the user triggering the natural language query information and its corresponding sentiment intensity; the target sentiment may include positive sentiment, negative sentiment, or neutral sentiment;

[0117] The intention recognition result, target entity information, target emotion and its corresponding emotion intensity are determined as the second information analysis result corresponding to the first information analysis result.

[0118] In this optional embodiment, the entity information parsing is also used to extract other entity information corresponding to the user, and the other entity information includes user ID, product name corresponding to the product currently queried by the user, manufacturer name corresponding to the product name, geographic information, time information and other information, wherein the geographic information can be information corresponding to the product shipping place or production place; the time information can be product production time or supply time / shipping time and other information.

[0119] It can be seen that in this optional embodiment, a semantic analysis model is used to perform multi-dimensional intent recognition on the first information analysis result, and a structured result containing intent categories and confidence scores is output. The setting of this mechanism can improve the accuracy of locating the core demands of user queries; it can also perform entity information extraction on the preprocessing results through the semantic analysis model, automatically identify and output target entity keywords that are highly relevant to the query, and achieve efficient conversion from unstructured text to structured data, thereby improving the efficiency and accuracy of subsequent knowledge base retrieval; and it can introduce an analysis module for emotional tendencies, accurately identify the emotional tendencies implied in the user's current query, and quantify the corresponding emotional intensity values, thereby achieving accurate perception of the user's emotional state, thereby providing data support for subsequent personalized and humanized response output.

[0120] In this optional embodiment, further, the above-mentioned method of performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result specifically includes:

[0121] Performing a first entity resolution on the first information analysis result according to the semantic analysis model to obtain a first entity resolution result corresponding to the first information analysis result; the first entity resolution is used to perform preliminary entity extraction on the first information analysis result;

[0122] Determining whether the first entity resolution result satisfies a disambiguation fusion condition; if so, performing a disambiguation fusion processing operation on the first entity resolution result to obtain a disambiguation fusion processing result corresponding to the first entity resolution result; the disambiguation fusion processing operation is used to disambiguate ambiguous entities and / or to fuse different entities with the same meaning; for example, for the entity "apple", determining whether it refers to fruit or Apple Inc. based on the context; for another example, identifying "Apple Inc." and "iPhone" as entities related to the same topic and associating them;

[0123] According to the semantic analysis model, a second entity resolution is performed on the disambiguation fusion processing result to obtain a second entity resolution result corresponding to the disambiguation fusion processing result as the target entity information for the first information analysis result; the second entity resolution includes key entity screening and entity information organization.

[0124] For example, if the user is querying product information, the key entity filter in this setting is used to adjust the focus to entities such as product name, brand name, and model number;

[0125] Optionally, after obtaining a second entity resolution result corresponding to the disambiguation fusion processing result, performing sampling verification on the second entity resolution result according to the semantic analysis model to obtain a sampling verification result corresponding to the second entity resolution result, the sampling verification result including entity recognition accuracy and completeness;

[0126] When the sampling verification result indicates that the entity recognition accuracy is lower than a preset accuracy threshold and / or the completeness is lower than a preset completeness threshold, model optimization is performed on the semantic analysis model according to the sampling verification result.

[0127] It can be seen that in this optional embodiment, by performing multi-level parsing operations on the first information analysis results, including preliminary entity extraction, disambiguation fusion, and deep key entity extraction and organization, the fineness and accuracy of the entity information parsing are improved.

[0128] In yet another optional embodiment, the method of performing the third information analysis operation on the second information analysis result according to the knowledge base to obtain the third information analysis result corresponding to the second information analysis result specifically includes:

[0129] Performing a first-level matching operation on the second information analysis result according to the knowledge base to obtain a first-level matching result corresponding to the second information analysis result; the first-level matching operation includes knowledge item retrieval; the first-level matching result includes at least one knowledge item to be screened;

[0130] According to the knowledge base, a secondary matching operation is performed on the primary matching results to obtain a secondary matching result corresponding to the primary matching result; the secondary matching operation includes in-depth analysis based on the first information analysis result and the second information analysis result, matching degree calculation, applicability scoring and sorting; the secondary matching result includes target knowledge items that meet the screening requirements;

[0131] According to the second information analysis result, information filling and emotion adjustment are performed on the secondary matching result to obtain the target output result as the third information analysis result.

[0132] In this optional embodiment, the first-level matching operation may include knowledge item retrieval and / or response template retrieval; correspondingly, the matching result includes a knowledge item to be screened and / or a response template to be screened.

[0133] It can be seen that in this optional embodiment, a secondary matching process is designed. By integrating the first information analysis results (context features) and the second information analysis results (semantic features), combined with multi-dimensional analysis (including matching calculation, applicability scoring and dynamic sorting), the knowledge items are screened step by step and accurately. This solves the problem of response deviation caused by a single matching dimension, which is conducive to improving the accuracy and applicability of the target knowledge items screened; and, for the secondary matching results, a corresponding information filling and emotion adjustment mechanism is further set up to achieve intelligent adjustment between standardized knowledge items and personalized response content, solving the problems of low matching efficiency and mechanized response in traditional knowledge base applications, and improving the humanization of the output third information analysis results / overall intelligent customer system, which is conducive to improving the user's interactive experience.

[0134] In another optional embodiment, the above-mentioned method of performing a secondary matching operation on the primary matching result according to the knowledge base to obtain the secondary matching result corresponding to the primary matching result specifically includes:

[0135] For each knowledge item to be screened, performing a text similarity calculation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a similarity calculation result between the second information analysis result and the knowledge item to be screened; the text similarity calculation operation includes a text preprocessing operation, a feature extraction operation and a similarity calculation operation;

[0136] Performing a semantic annotation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a semantic annotation result corresponding to the second information analysis result and the knowledge item to be screened; the semantic annotation operation includes a syntactic structure parsing operation, a semantic role annotation operation, and a semantic role comparison operation;

[0137] Based on the similarity calculation result and the semantic annotation result corresponding to the knowledge item to be screened, a comprehensive similarity score between the second information analysis result and the knowledge item to be screened is calculated, and a suitability score corresponding to the knowledge item to be screened and the user's query requirements and / or conversation scenario is calculated;

[0138] Based on the comprehensive similarity score and applicability score corresponding to each knowledge item to be screened, all knowledge items to be screened are scored and sorted, and any knowledge item to be screened that is ranked within the preset ranking range is selected from the corresponding score sorting results and recorded as the target knowledge item, and used as the secondary matching result.

[0139] In this optional embodiment, the text preprocessing operation includes word segmentation, stop word removal, stemming, and phrase reorganization; the text preprocessing is used to process the second information analysis result into a standardized text representation;

[0140] The semantic role labeling operation is used to label all semantic roles corresponding to the second information analysis result, and all semantic roles include agent role, patient role, tool, time and place.

[0141] In this optional embodiment, the agent role is the executing entity of the matter, such as the purchaser; the recipient role is the executed entity of the matter, such as the supplier who wins the bet with the purchaser.

[0142] In this optional embodiment, the text similarity calculation operation can be implemented based on a pre-built deep learning model. Furthermore, the deep learning model is specifically designed for multi-dimensional text similarity calculation and evaluation; the deep learning model includes a standardized preprocessing process based on word segmentation, stop word removal, stemming, and phrase reorganization, and also involves conventional text feature extraction and similarity calculation. Furthermore, through this deep learning model, it is possible to effectively eliminate differences in natural language expression, accurately quantify the semantic proximity and similarity between the second information analysis results and the knowledge items, and provide a data basis for subsequent knowledge item screening.

[0143] It can be seen that in this optional embodiment, by constructing a three-level screening mechanism corresponding to "text similarity-semantic role-two-dimensional scoring", it is possible to quickly locate the optimal response content from massive knowledge entries, effectively solving the semantic gap and low scenario adaptability problems existing in traditional knowledge matching, improving the service accuracy and adaptability of the B2B intelligent customer service system, and at the same time improving the user's interactive experience and business processing efficiency of the B2B intelligent customer service system.

[0144] Example 3

[0145] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of a B2B system intelligent customer service management and control system based on natural language processing disclosed in an embodiment of the present invention. Figure 3 As shown, the B2B system intelligent customer service management and control system based on natural language processing may include an acquisition module 301, a pre-processing module 302 and a target processing module 303, wherein:

[0146] The acquisition module 301 is used to acquire natural language query information fed back by users through multiple channels. The natural language query information includes query information triggered by users through the B2B platform.

[0147] The preprocessing module 302 is used to perform a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data.

[0148] The target processing module 303 is used to perform a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to natural language query information.

[0149] It can be seen that implementation Figure 3 The described B2B system intelligent customer service management and control system based on natural language processing performs pre-processing on natural language query information and converts it into standardized and structured text data, which can effectively remove redundant data and improve the processing efficiency and accuracy of the natural language query information; and, by combining the context understanding module, semantic analysis model and knowledge base, conducts multi-dimensional and in-depth information analysis on the pre-processing results, so as to accurately capture the user's query intention and explore the user's potential needs, which is conducive to improving the accuracy and reliability of subsequent responses to natural language query information triggered by the user and improving the user's query experience.

[0150] In an alternative embodiment, see Figure 4 , Figure 4 This is a structural diagram of another B2B system intelligent customer service management and control system based on natural language processing disclosed in an embodiment of the present invention. Figure 4 As shown, the target processing module 303 includes:

[0151] A first information analysis submodule 3031 is configured to perform a first information analysis operation on the preprocessing result according to the context understanding module to obtain a first information analysis result corresponding to the preprocessing result; the first information analysis operation includes at least one of history record query, session state analysis, and user preference analysis;

[0152] The second information analysis submodule 3032 is configured to perform a second information analysis operation on the first information analysis result according to the semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result; the second information analysis operation includes at least one of user query intent recognition, entity information parsing, and user sentiment tendency analysis;

[0153] The third information analysis submodule 3033 is used to perform a third information analysis operation on the second information analysis result according to the knowledge base, and obtain a third information analysis result corresponding to the second information analysis result as the target processing result; the third information analysis operation is used to match the response content of the response to the second information analysis result.

[0154] It can be seen that in this optional embodiment, through the multi-level information analysis and processing process, the B2B system can achieve efficient and accurate responses to user queries. Specifically, it can use the context understanding module, combined with historical record query, session state analysis and user preference analysis, to comprehensively capture the context information of user queries, which is conducive to improving the correlation between information analysis and context and thus improving the accuracy of information analysis; and it can use the semantic analysis model to deeply identify and analyze user query intentions, entity information and emotional tendencies, which is conducive to improving the accuracy of analysis of users' real query needs; it can also intelligently match the analysis results based on the knowledge base, quickly locate and generate response content that is highly consistent with user needs, realize the efficient use of knowledge resources, and improve the intelligence level of the response process corresponding to user query needs and improve service efficiency and quality.

[0155] In another optional embodiment, the first information analysis submodule 3031 performs a first information analysis operation on the preprocessing result according to the context understanding module, and obtains the first information analysis result corresponding to the preprocessing result in a manner specifically including:

[0156] According to the preset session state model, real-time monitoring of the session state after the user triggers the natural language query information, the session state includes any of the states of session start, ongoing, paused, and ended;

[0157] When the session state is determined to be a session start or in progress, detecting whether there is a historical conversation record corresponding to the preprocessing result; when the historical conversation record is detected to exist, performing information extraction on the historical conversation record according to the context understanding module to obtain a record summary corresponding to the historical conversation record, the record summary including at least one of the historical conversation content, the question type, and the solution, the correlation degree of which to the natural language query information is higher than a preset correlation threshold;

[0158] The context understanding module performs user preference analysis on historical conversation records to obtain the user's query preference items, potential demand items, and points of interest items as user preference information;

[0159] The record summary and user preference information are associated and integrated with the preprocessing result to update the preprocessing result and obtain a first information analysis result.

[0160] It can be seen that in this optional embodiment, based on the preset session state model, real-time monitoring of the entire cycle of the user session can be achieved, and key nodes such as the start, progress, pause, and end of the session can be accurately identified, providing a dynamic context benchmark for subsequent information analysis, ensuring that the analysis process is highly synchronized with the session process; and, in the active phase of the session (start / progress), it can automatically retrieve and associate historical conversation records, and extract highly correlated summary information (including historical conversation content, problem type, solution, etc.) through the context understanding module, thereby effectively reusing historical interaction experience and improving the efficiency of current queries; and, it can also perform multi-dimensional preference analysis based on historical conversation records, accurately construct user query preferences, potential needs and interest point portraits, realize in-depth enhancement of personalized service capabilities, and further improve the matching degree and accuracy of response content with the user's real needs.

[0161] In another optional embodiment, the second information analysis submodule 3032 performs a second information analysis operation on the first information analysis result according to the semantic analysis model, and obtains the second information analysis result corresponding to the first information analysis result in a manner specifically including:

[0162] Performing user query intent recognition on the first information analysis result according to the semantic analysis model to obtain a user intent recognition result; the intent recognition result includes an intent category and a confidence score;

[0163] Performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result; the entity information parsing is used to extract keywords corresponding to the information query in the preprocessing result; the target entity information includes one or more entity keywords;

[0164] Performing sentiment analysis on the first information analysis result according to the semantic analysis model to obtain the target sentiment of the user triggering the natural language query information and its corresponding sentiment intensity; the target sentiment may include positive sentiment, negative sentiment, or neutral sentiment;

[0165] The intention recognition result, target entity information, target emotion and its corresponding emotion intensity are determined as the second information analysis result corresponding to the first information analysis result.

[0166] It can be seen that in this optional embodiment, a semantic analysis model is used to perform multi-dimensional intent recognition on the first information analysis result, and a structured result containing intent categories and confidence scores is output. The setting of this mechanism can improve the accuracy of locating the core demands of user queries; it can also perform entity information extraction on the preprocessing results through the semantic analysis model, automatically identify and output target entity keywords that are highly relevant to the query, and achieve efficient conversion from unstructured text to structured data, thereby improving the efficiency and accuracy of subsequent knowledge base retrieval; and it can introduce an analysis module for emotional tendencies, accurately identify the emotional tendencies implied in the user's current query, and quantify the corresponding emotional intensity values, thereby achieving accurate perception of the user's emotional state, thereby providing data support for subsequent personalized and humanized response output.

[0167] In another optional embodiment, the third information analysis submodule 3033 performs a third information analysis operation on the second information analysis result according to the knowledge base, and obtains the third information analysis result corresponding to the second information analysis result in a manner specifically including:

[0168] Performing a first-level matching operation on the second information analysis result according to the knowledge base to obtain a first-level matching result corresponding to the second information analysis result; the first-level matching operation includes knowledge item retrieval; the first-level matching result includes at least one knowledge item to be screened;

[0169] According to the knowledge base, a secondary matching operation is performed on the primary matching results to obtain a secondary matching result corresponding to the primary matching result; the secondary matching operation includes in-depth analysis based on the first information analysis result and the second information analysis result, matching degree calculation, applicability scoring and sorting; the secondary matching result includes target knowledge items that meet the screening requirements;

[0170] According to the second information analysis result, information filling and emotion adjustment are performed on the secondary matching result to obtain the target output result as the third information analysis result.

[0171] It can be seen that in this optional embodiment, a secondary matching process is designed. By integrating the first information analysis results (context features) and the second information analysis results (semantic features), combined with multi-dimensional analysis (including matching calculation, applicability scoring and dynamic sorting), the knowledge items are screened step by step and accurately. This solves the problem of response deviation caused by a single matching dimension, which is conducive to improving the accuracy and applicability of the target knowledge items screened; and, for the secondary matching results, a corresponding information filling and emotion adjustment mechanism is further set up to achieve intelligent adjustment between standardized knowledge items and personalized response content, solving the problems of low matching efficiency and mechanized response in traditional knowledge base applications, and improving the humanization of the output third information analysis results / overall intelligent customer system, which is conducive to improving the user's interactive experience.

[0172] In another optional embodiment, the third information analysis submodule 3033 performs a secondary matching operation on the primary matching result according to the knowledge base, and obtains the secondary matching result corresponding to the primary matching result in the following manner:

[0173] For each knowledge item to be screened, performing a text similarity calculation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a similarity calculation result between the second information analysis result and the knowledge item to be screened; the text similarity calculation operation includes a text preprocessing operation, a feature extraction operation and a similarity calculation operation;

[0174] Performing a semantic annotation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a semantic annotation result corresponding to the second information analysis result and the knowledge item to be screened; the semantic annotation operation includes a syntactic structure parsing operation, a semantic role annotation operation, and a semantic role comparison operation;

[0175] Based on the similarity calculation result and the semantic annotation result corresponding to the knowledge item to be screened, a comprehensive similarity score between the second information analysis result and the knowledge item to be screened is calculated, and a suitability score corresponding to the knowledge item to be screened and the user's query requirements and / or conversation scenario is calculated;

[0176] Based on the comprehensive similarity score and applicability score corresponding to each knowledge item to be screened, all knowledge items to be screened are scored and sorted, and any knowledge item to be screened that is ranked within the preset ranking range is selected from the corresponding score sorting results and recorded as the target knowledge item, and used as the secondary matching result.

[0177] In this optional embodiment, the text preprocessing operation includes word segmentation, stop word removal, stemming, and phrase reorganization; the text preprocessing is used to process the second information analysis result into a standardized text representation;

[0178] The semantic role labeling operation is used to label all semantic roles corresponding to the second information analysis result, and all semantic roles include agent role, patient role, tool, time and place.

[0179] It can be seen that in this optional embodiment, by constructing a three-level screening mechanism corresponding to "text similarity-semantic role-two-dimensional scoring", it is possible to quickly locate the optimal response content from massive knowledge entries, effectively solving the semantic gap and low scenario adaptability problems existing in traditional knowledge matching, improving the service accuracy and adaptability of the B2B intelligent customer service system, and at the same time improving the user's interactive experience and business processing efficiency of the B2B intelligent customer service system.

[0180] Example 4

[0181] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of a B2B system intelligent customer service management and control device based on natural language processing disclosed in an embodiment of the present invention. Figure 5 As shown, the B2B system intelligent customer service management and control device based on natural language processing may include:

[0182] A memory 401 storing executable program code;

[0183] a processor 402 coupled to the memory 401;

[0184] The processor 402 calls the executable program code stored in the memory 401 to execute the steps of the B2B system intelligent customer service management and control method based on natural language processing described in the first embodiment of the present invention or the second embodiment of the present invention.

[0185] Example 5

[0186] An embodiment of the present invention discloses a computer storage medium storing computer instructions. When the computer instructions are called, they are used to execute the steps of the B2B system intelligent customer service management and control method based on natural language processing described in Embodiment 1 or Embodiment 2 of the present invention.

[0187] Example 6

[0188] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer storage medium storing a computer program, and the computer program is operable to cause a computer to execute the steps of the B2B system intelligent customer service management and control method based on natural language processing described in Example 1 or Example 2.

[0189] The system / device embodiments described above are merely illustrative. Modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical modules, i.e., they may be located in one place or distributed across multiple network modules. Some or all of these modules may be selected based on actual needs to achieve the objectives of the present embodiment. Those skilled in the art will be able to understand and implement the present invention without inventive effort.

[0190] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus the necessary general hardware platform, or of course, by means of hardware. Based on this understanding, the above technical solution, in essence, or the portion that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer storage medium, including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0191] Finally, it should be noted that the B2B system intelligent customer service management method and system based on natural language processing disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, and are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features therein may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A B2B system intelligent customer service management and control method based on natural language processing, characterized in that: The method comprises: Obtaining natural language query information fed back by users through multiple channels, wherein the natural language query information includes query information triggered by the users through the B2B platform; Performing a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is used to convert the natural language query information into recognizable text data; A target processing operation is performed on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to the natural language query information.

2. The B2B system intelligent customer service management and control method based on natural language processing according to claim 1 is characterized in that: The performing a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result includes: performing a first information analysis operation on the preprocessing result according to the context understanding module to obtain a first information analysis result corresponding to the preprocessing result; the first information analysis operation includes at least one of a history query, a session state analysis, and a user preference analysis; performing a second information analysis operation on the first information analysis result according to the semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result; the second information analysis operation includes at least one of user query intent recognition, entity information parsing, and user sentiment tendency analysis; According to the knowledge base, a third information analysis operation is performed on the second information analysis result to obtain a third information analysis result corresponding to the second information analysis result as the target processing result; the third information analysis operation is used to match the response content in response to the second information analysis result.

3. The B2B system intelligent customer service management and control method based on natural language processing according to claim 2 is characterized in that: The context understanding module performs a first information analysis operation on the preprocessing result to obtain a first information analysis result corresponding to the preprocessing result, including: According to a preset session state model, real-time monitoring of the session state after the user triggers the natural language query information, wherein the session state includes any one of the states of session start, ongoing, paused, and ended; When it is determined that the session state is the session start or the session in progress, detecting whether there is a historical conversation record corresponding to the preprocessing result; and when the historical conversation record is detected to exist, performing information extraction on the historical conversation record according to the context understanding module to obtain a record summary corresponding to the historical conversation record, the record summary including at least one of the historical conversation content, the question type, and the solution, the correlation degree of which to the natural language query information is higher than a preset correlation threshold; Performing user preference analysis on the historical conversation records according to the context understanding module to obtain the user's query preference items, potential demand items, and points of interest items as user preference information; The record summary and the user preference information are associated and integrated with the preprocessing result to update the preprocessing result and obtain a first information analysis result.

4. The B2B system intelligent customer service management and control method based on natural language processing according to claim 2 or 3, characterized in that: The performing a second information analysis operation on the first information analysis result according to the semantic analysis model to obtain a second information analysis result corresponding to the first information analysis result includes: Performing user query intent recognition on the first information analysis result according to the semantic analysis model to obtain the user's intent recognition result; the intent recognition result includes an intent category and a confidence score; performing entity information parsing on the first information analysis result according to the semantic analysis model to obtain target entity information for the first information analysis result; the entity information parsing is used to extract keywords corresponding to the information query in the preprocessing result; the target entity information includes one or more entity keywords; Performing sentiment analysis on the first information analysis result according to the semantic analysis model to obtain the target sentiment of the user triggering the natural language query information and its corresponding sentiment intensity; the target sentiment may include positive sentiment, negative sentiment, or neutral sentiment; The intention recognition result, the target entity information, the target emotion and its corresponding emotion intensity are determined as a second information analysis result corresponding to the first information analysis result.

5. The B2B system intelligent customer service management and control method based on natural language processing according to claim 2 or 3, characterized in that: The performing a third information analysis operation on the second information analysis result according to the knowledge base to obtain a third information analysis result corresponding to the second information analysis result includes: performing a first-level matching operation on the second information analysis result according to the knowledge base to obtain a first-level matching result corresponding to the second information analysis result; the first-level matching operation includes knowledge item retrieval; the first-level matching result includes at least one knowledge item to be screened; performing a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result; the secondary matching operation includes in-depth analysis, matching degree calculation, applicability scoring and ranking based on the first information analysis result and the second information analysis result; the secondary matching result includes target knowledge items that meet the screening requirements; According to the second information analysis result, information filling and emotion adjustment are performed on the secondary matching result to obtain a target output result as the third information analysis result.

6. The B2B system intelligent customer service management and control method based on natural language processing according to claim 5 is characterized in that: The step of performing a secondary matching operation on the primary matching result according to the knowledge base to obtain a secondary matching result corresponding to the primary matching result includes: For each of the knowledge items to be screened, performing a text similarity calculation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a similarity calculation result between the second information analysis result and the knowledge item to be screened; the text similarity calculation operation includes a text preprocessing operation, a feature extraction operation, and a similarity calculation operation; Performing a semantic annotation operation on the second information analysis result and the knowledge item to be screened according to the knowledge base to obtain a semantic annotation result corresponding to the second information analysis result and the knowledge item to be screened; the semantic annotation operation includes a syntactic structure parsing operation, a semantic role annotation operation, and a semantic role comparison operation; Based on the similarity calculation result and the semantic annotation result corresponding to the knowledge item to be screened, a comprehensive similarity score between the second information analysis result and the knowledge item to be screened is calculated, and a suitability score corresponding to the knowledge item to be screened and the user's query requirements and / or conversation scenario is calculated; Based on the comprehensive similarity score and the applicability score corresponding to each of the knowledge items to be screened, all the knowledge items to be screened are scored and sorted, and any of the knowledge items to be screened that is ranked within the preset ranking range is selected from the corresponding scoring and sorting results, recorded as the target knowledge item, and used as the secondary matching result.

7. The B2B system intelligent customer service management and control method based on natural language processing according to claim 6, characterized in that: The text preprocessing operations include word segmentation, stop word removal, stemming, and phrase reorganization; the text preprocessing is used to process the second information analysis result into a standardized text representation; The semantic role labeling operation is used to label all semantic roles corresponding to the second information analysis result, and all the semantic roles include agent role, patient role, tool, time and place.

8. A B2B system intelligent customer service management and control system based on natural language processing, characterized by: The system comprises: An acquisition module, configured to acquire natural language query information fed back by users through multiple channels, wherein the natural language query information includes query information triggered by the users through the B2B platform; A preprocessing module, configured to perform a preprocessing operation on the natural language query information to obtain a preprocessing result for the natural language query information; the preprocessing is configured to convert the natural language query information into recognizable text data; A target processing module is used to perform a target processing operation on the preprocessing result to obtain a target processing result for the preprocessing result; the target processing operation includes at least one of a first information analysis operation based on a context understanding module, a second information analysis operation based on a semantic analysis model, and a third information analysis operation based on a knowledge base; the target processing result is used to indicate a response item in response to the natural language query information.

9. A B2B system intelligent customer service management and control device based on natural language processing, characterized in that: The device comprises: a memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the B2B system intelligent customer service management and control method based on natural language processing according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that The computer storage medium stores computer instructions, which, when called, are used to execute the B2B system intelligent customer service management and control method based on natural language processing according to any one of claims 1 to 7.

Images