Consultation content processing method and equipment based on intelligent customer service and storage medium

Abstract

The invention provides a consultation content processing method and device based on intelligent customer service and a storage medium. The method comprises the following steps: receiving a plurality of to-be-processed consultation contents and consultation time sent by a target client of a target user; obtaining associated consultation content according to the consultation time; performing feature analysis on the associated consultation content according to preset analysis information to obtain a first content feature; matching the first content feature by adopting a preset knowledge base to obtain target reply information; obtaining pre-consultation content according to the consultation time; extracting a second content feature from the front consultation content; calculating the content similarity between the first content feature and the second content feature; screening out predicted consultation contents from the second content features according to the content similarity; sending the target reply information and the predicted consultation content to the target client, so that the target client displays the target reply information and the predicted consultation content; the problem of poor user experience caused by the scheme in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a consultation content processing method, device and storage medium based on intelligent customer service. Background Art

[0002] Currently, with the development of artificial intelligence technology, intelligent customer service based on human-computer dialogue technology is widely used in scenarios for answering user questions.

[0003] In the prior art, when intelligent customer service interacts with users through question-and-answer sessions, it extracts corresponding keywords based on the questions sent by the users, and then matches relevant reference content in the knowledge base based on the extracted keywords, and then the intelligent customer service replies to the users with the matched reference content.

[0004] However, in the solutions of the prior art, the intelligent customer service cannot associate the text content input by the user twice, and further infer the user's intention, and thus cannot accurately reply to the user with relevant reply information. Furthermore, based on the inability to accurately reply to the user with reply information, it is also impossible to provide the user with corresponding predictive consulting information. Therefore, the solutions of the prior art have the problem of poor user experience. Summary of the invention

[0005] The present application provides a consultation content processing method, device and storage medium based on intelligent customer service, which are used to solve the problem of poor user experience caused by the prior art.

[0006] In the first aspect, the present application provides a consultation content processing method based on intelligent customer service, which is applied to a service device, including: receiving multiple consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry a consultation time; associating and merging consultation contents belonging to the same dialogue cycle according to the consultation time to obtain associated consultation contents; performing feature analysis on the associated consultation contents according to preset analysis information to obtain one or more first content features corresponding to the associated consultation contents; matching the one or more first content features with a preset knowledge base to obtain target reply information; obtaining pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation time. consulting content sent by other users through the client within a previously preset time interval; extracting one or more second content features from the previous consulting content; calculating the similarity between each of the first content features and each of the second content features to obtain content similarity; filtering out the predicted consulting content of the target user from the second content features according to the content similarity; sending the target reply information and the predicted consulting content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consulting content, and displays a first interactive control for sending the predicted consulting content, and / or a second interactive control for secondary editing of the predicted consulting content.

[0007] In one possible design, the related consulting content is feature parsed according to preset parsing information to obtain one or more first content features corresponding to the related consulting content, including: performing a natural language description of the related consulting content according to the preset parsing information to obtain target parsing information; and performing semantic analysis on the target parsing information by a pre-trained large language model to obtain the first content features.

[0008] In a possible design, the related consulting content is described in natural language according to the preset resolution information to obtain target resolution information, including: identifying business characteristics of the related consulting content to obtain business characteristic identifiers; matching the business characteristic identifiers with the preset resolution information to determine characteristic resolution information; and describing the related consulting content in natural language according to the characteristic resolution information to obtain the target resolution information.

[0009] In one possible design, the consulting contents belonging to the same conversation cycle are associated and merged according to the consultation time to obtain associated consulting contents, including: associating and merging the consulting contents belonging to the same conversation cycle according to the consultation time to obtain initial consulting contents; obtaining historical consulting contents; wherein the historical consulting contents are consulting contents sent by clients of other users before the consultation time, and the historical consulting contents carry historical consultation time; according to the time correlation between the consultation time and the historical consultation time, one or more reference consulting contents are extracted from the historical consulting contents; and the initial consulting contents are respectively associated with the one or more reference consulting contents to obtain the associated consulting contents.

[0010] In a possible design, one or more reference consultation contents are extracted from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time, including: dividing the historical consultation time into intervals according to a preset start time and a preset end time to obtain a target time interval; obtaining a target time identifier according to a relative time position relationship of the consultation time within the target time interval; dividing the historical consultation contents into consultation contents according to the target time interval to obtain interval consultation contents; and extracting the reference consultation contents from the interval consultation contents according to the target time identifier.

[0011] In a possible design, the initial consulting content is respectively associated with the one or more reference consulting contents to obtain the associated consulting contents, including: dividing the consultation time into a first consultation time interval and a second consultation time interval according to the chronological order; performing sentence component analysis on the initial consulting content based on the first consultation time interval to obtain at least one first keyword; performing sentence component analysis on the initial consulting content based on the second consultation time interval to obtain at least one second keyword; and performing association processing on the one or more reference consulting contents according to the at least one first keyword and the at least one second keyword to obtain the associated consulting contents.

[0012] In a possible design, the target time interval includes a weekly time interval, a monthly time interval, a quarterly time interval, and an annual time interval.

[0013] In a possible design, the calculating the similarity between each of the first content features and each of the second content features to obtain the content similarity includes: calculating the user similarity between the target user and the other users; screening each of the second content features according to the user similarity to obtain at least two target second content features; and calculating the similarity between each of the first content features and each of the target second content features to obtain the content similarity.

[0014] In one possible design, the method also includes: performing sentence splitting on the predictive consultation content to obtain one or more predicted candidate words; sending the one or more predicted candidate words to the target client of the target user, so that the target client of the target user displays the one or more predicted candidate words, and displays a third interactive control for the target user to select the one or more predicted candidate words.

[0015] In a second aspect, the present application provides a consultation content processing device based on intelligent customer service, comprising:

[0016] A receiving module, configured to receive a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry a consultation time;

[0017] A first processing module is used to associate and merge consultation contents belonging to the same dialogue cycle according to the consultation time to obtain associated consultation contents; perform feature analysis on the associated consultation contents according to preset analysis information to obtain one or more first content features corresponding to the associated consultation contents; and use a preset knowledge base to match the one or more first content features to obtain target reply information;

[0018] The second processing module is used to obtain the pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time; extract one or more second content features from the pre-consultation content; calculate the similarity between each of the first content features and each of the second content features to obtain content similarity; and filter out the predicted consultation content of the target user from the second content features according to the content similarity;

[0019] An output module is used to send the target reply information and the predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consultation content, and displays a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content.

[0020] In a possible design, when the first processing module performs feature analysis on the related consulting content according to preset analysis information to obtain one or more first content features corresponding to the related consulting content, it is specifically used to: perform a natural language description of the related consulting content according to the preset analysis information to obtain target analysis information; and perform semantic analysis on the target analysis information using a pre-trained large language model to obtain the first content features.

[0021] In a possible design, when the first processing module performs a natural language description of the related consulting content according to the preset resolution information to obtain the target resolution information, it is specifically used to: identify the business characteristics of the related consulting content to obtain a business characteristic identifier; match the business characteristic identifier with the preset resolution information to determine the characteristic resolution information; and perform a natural language description of the related consulting content according to the characteristic resolution information to obtain the target resolution information.

[0022] In one possible design, when the first processing module associates and merges the consulting contents belonging to the same conversation cycle according to the consultation time to obtain associated consulting contents, it is specifically used to: associate and merge the consulting contents belonging to the same conversation cycle according to the consultation time to obtain initial consulting contents; obtain historical consulting contents; wherein the historical consulting contents are consulting contents sent by the clients of other users before the consultation time, and the historical consulting contents carry the historical consultation time; based on the time correlation between the consultation time and the historical consultation time, extract one or more reference consulting contents from the historical consulting contents; and associate the initial consulting contents respectively through the one or more reference consulting contents to obtain the associated consulting contents.

[0023] In a possible design, when the first processing module extracts one or more reference consultation contents from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time, it is specifically used to: divide the historical consultation time into intervals according to a preset start time and a preset end time to obtain a target time interval; obtain a target time identifier according to the relative time position relationship of the consultation time within the target time interval; divide the historical consultation contents into consultation contents according to the target time interval to obtain interval consultation contents; and extract the reference consultation contents from the interval consultation contents according to the target time identifier.

[0024] In a possible design, when the first processing module associates the initial consulting content with the one or more reference consulting contents respectively to obtain the associated consulting content, it is specifically used to: divide the consultation time into a first consultation time interval and a second consultation time interval according to the chronological order; based on the first consultation time interval, perform sentence component analysis on the initial consulting content to obtain at least one first keyword; based on the second consultation time interval, perform sentence component analysis on the initial consulting content to obtain at least one second keyword; and based on the at least one first keyword and the at least one second keyword, perform association processing on the one or more reference consulting contents respectively to obtain the associated consulting content.

[0025] In a possible design, the target time interval includes a weekly time interval, a monthly time interval, a quarterly time interval, and an annual time interval.

[0026] In a possible design, when the second processing module calculates the similarity between each of the first content features and each of the second content features to obtain the content similarity, it is specifically used to: calculate the user similarity between the target user and the other users; filter each of the second content features according to the user similarity to obtain at least two target second content features; calculate the similarity between each of the first content features and each of the target second content features to obtain the content similarity.

[0027] In a possible design, the consulting content processing device based on intelligent customer service is also used to: perform sentence splitting on the predicted consulting content to obtain one or more predicted candidate words; sending the one or more predicted candidate words to the target client of the target user, so that the target client of the target user displays the one or more predicted candidate words, and displays a third interactive control for the target user to select the one or more predicted candidate words.

[0028] In a third aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor and a memory; the memory stores computer execution instructions; the at least one processor executes the computer execution instructions stored in the memory, so that the at least one processor executes the consultation content processing method based on intelligent customer service as described in the first aspect and various possible designs of the first aspect.

[0029] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer execution instructions are stored. When a processor executes the computer execution instructions, the consultation content processing method based on intelligent customer service as described in the first aspect and various possible designs of the first aspect is implemented.

[0030] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the consultation content processing method based on intelligent customer service as described in the first aspect and various possible designs of the first aspect.

[0031] The consultation content processing method, device and storage medium based on intelligent customer service provided in the present application receive multiple consultation contents to be processed sent by the target client of the target user, wherein the consultation contents carry the consultation time; according to the consultation time, the consultation contents belonging to the same dialogue cycle are associated and merged to obtain associated consultation contents; according to the preset analysis information, the associated consultation contents are feature analyzed to obtain one or more first content features corresponding to the associated consultation contents; the one or more first content features are matched with the preset knowledge base to obtain the target reply information; according to the consultation time, the pre-consultation content is obtained, wherein the pre-consultation content is the pre-consultation content before the consultation time. Assume that within a time interval, other users send consulting content through the client; extract one or more second content features from the previous consulting content; calculate the similarity between each of the first content features and each of the second content features to obtain content similarity; filter out the predicted consulting content of the target user from the second content features according to the content similarity; send the target reply information and the predicted consulting content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consulting content, and displays a first interactive control for sending the predicted consulting content, and / or a second interactive control for secondary editing of the predicted consulting content. By associating and merging the consultation contents according to the consultation time of the target user, the associated consultation contents are obtained; and then, in combination with the resolution identifier corresponding to the preset resolution information, the associated consultation contents are feature analyzed to obtain the first content feature; further, the first content feature is matched with the preset knowledge base, so that the target reply information of the intelligent customer service to the target user can be determined from the preset knowledge base; further, by extracting the second content feature from the pre-consultation content and calculating the similarity between the first content feature and the second content feature, the predicted consultation content of the target user is screened out from the second content feature, and the target user is provided with a "direct send" interaction component and / or a "secondary editing" interaction component for the second content feature; the solution of the embodiment of the present application not only solves the problem that the intelligent customer service cannot accurately reply to the user with relevant reply information, but also solves the problem that the corresponding predicted consultation information cannot be provided to the user, thereby solving the problem of poor user experience caused by the solution of the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0033] Figure 1 An application scenario diagram of the consultation content processing method based on intelligent customer service provided in an embodiment of the present application;

[0034] Figure 2 A flowchart of a method for processing consultation content based on intelligent customer service provided in one embodiment of the present application;

[0035] Figure 3 for Figure 2 A schematic diagram of specific implementation steps of step S102 in the illustrated embodiment;

[0036] Figure 4 for Figure 3 A schematic diagram of specific implementation steps of step S1021 in the illustrated embodiment;

[0037] Figure 5 A schematic diagram of a process for determining a first content feature based on a service feature identifier provided in this application;

[0038] Figure 6 A flowchart of a method for processing consultation content based on intelligent customer service provided in another embodiment of the present application;

[0039] Figure 7 A schematic diagram of reference consultation content determined based on different consultation times provided for this application;

[0040] Figure 8 for Figure 6 A schematic diagram of specific implementation steps of step S203 in the illustrated embodiment;

[0041] Fig. 9 A schematic diagram of determining reference consultation content based on interval consultation content provided by this application;

[0042] Fig.10 for Figure 6 A schematic diagram of specific implementation steps of step S204 in the illustrated embodiment;

[0043] Fig.11 A schematic diagram of the structure of a consultation content processing device based on intelligent customer service provided in one embodiment of the present application;

[0044] Fig.12 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0045] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0046] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0047] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of information such as financial data or user data involved shall comply with the provisions of relevant laws and regulations and shall not violate public order and good morals.

[0048] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0049] The application scenarios of the embodiments of the present application are explained below:

[0050] Figure 1 This is an application scenario diagram of the consultation content processing method based on intelligent customer service provided in the embodiment of the present application. The consultation content processing method based on intelligent customer service provided in the embodiment of the present application can be applied to the scenario where the intelligent customer service interacts with the user for question and answer. For example, Figure 1 As shown, the execution subject of the method provided in the embodiment of the present application can be an electronic control unit, a server or a terminal device. The server is used as the execution subject for explanation. The server of the application with the consultation content processing function based on intelligent customer service is deployed in the server, and the terminal device on the user side is deployed with the client of the application. The server and the terminal device communicate based on the server-client, and then obtain the consultation content input by the user sent by the terminal device; the server processes the consultation content input by the user through the consultation content processing method based on intelligent customer service provided by the present application, and then obtains the reply content for replying to the customer, thereby realizing the question-answer interaction between the intelligent customer service and the user.

[0051] Currently, when the intelligent customer service interacts with the user in a question-and-answer session, it will extract the corresponding keywords based on the questions sent by the user, and then match the relevant reference content in the knowledge base based on the extracted keywords, and then the intelligent customer service will reply to the user with the matched reference content. However, based on the user's personal habits, the content entered is short words. For example, the text content entered for the first time is "Product A_1", and the text content entered for the second time is "Product A_2". Based on the execution logic of extracting keywords, the intelligent customer service will only provide the user with relevant introductions to Product A_1 and Product A_2 respectively. The intelligent customer service cannot associate the text content entered by the user twice, and further infer the user's intention, and then provide the user with comparative information about Product A_1 and Product A_2. Furthermore, for example, the text content entered by the user for the first time is "Product A_ 1", and the second input text content is "buy", then the intelligent customer service will only provide the user with the relevant introduction of product A_1 and the word explanation of "buy" based on the execution logic of extracting keywords. Similarly, the intelligent customer service cannot associate the text content input by the user twice, and further infer the user's intention, and then provide the user with a purchase link or purchase suggestion for product A_1; further, on the basis of being unable to accurately reply to the user with the reference content, it is also impossible to provide the user with corresponding predictive consulting information and corresponding interactive controls. Therefore, the solution of the prior art leads to the problem of poor user experience.

[0052] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0053] Figure 2 A flowchart of a method for processing consultation content based on intelligent customer service is provided for one embodiment of the present application, such as Figure 2 As shown, the execution subject of the consultation content processing method based on intelligent customer service provided in this embodiment can be an electronic control unit, a server or a terminal device. By way of example, this embodiment is described with the server as the execution subject of the method of this embodiment. The consultation content processing method based on intelligent customer service provided in this embodiment includes the following steps:

[0054] Step S100, receiving a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry consultation time.

[0055] Step S101 , associating and merging consultation contents belonging to the same conversation cycle according to consultation time to obtain associated consultation contents.

[0056] Exemplarily, the consultation content to be processed is used to indicate the consultation content sent by the target user who interacts with the intelligent customer service through the target client for question and answer, and the content forms of the consultation content include text form, audio form and image form, and the number of consultation content items is two or more, wherein the consultation content entered by the user (user consultation information) and the user login information when the user logs in to the application with the consultation content processing function based on the intelligent customer service have been authorized by the user for storage, use, processing and processing by the method of this application; further, the consultation content carries the consultation time; and then, the server associates and merges the consultation content belonging to the same dialogue cycle based on the consultation time to obtain the associated consultation content. Specifically, for example, the text content entered by the user at time_1 is "Product A_1", and the text content entered at time_2 is "Purchase". If the time interval between time_1 and time_2 is less than the preset duration, the text content "Product A_1" and the text content "Purchase" are associated as the user's consulting content in one dialogue cycle, that is, the associated consulting content info_0 consisting of the text content "Product A_1" and the text content "Purchase" is obtained; if the time interval between time_1 and time_2 is greater than or equal to the preset duration, the text content "Product A_1" and the text content "Purchase" are respectively determined as the user's consulting content in two dialogue cycles, that is, the associated consulting content info_1 determined by the text content "Product A_1" and the associated consulting content info_2 determined by the text content "Purchase" are obtained.

[0057] When the consulting content sent by the user is in audio and / or image form, the server first converts the consulting content in audio and / or image form into text form before further processing.

[0058] Step S102: performing feature analysis on the associated consulting content according to preset analysis information to obtain one or more first content features corresponding to the associated consulting content.

[0059] Exemplarily, the preset parsing information is used to indicate the parsing identifier when parsing the associated consulting content. By parsing the associated consulting content with the parsing identifier corresponding to the preset parsing information, one or more first content features corresponding to the associated consulting content can be obtained. Specifically, for example, it can be known from the embodiment of step S101 that based on the time interval between the moment time_1 and the moment time_2 being less than the preset duration, the associated consulting content info_0 is composed of the text content "product A_1" and the text content "purchase". Accordingly, when parsing the associated consulting content info_0, it is necessary to match the corresponding parsing identifier label_1 for the case where the number of text contents is two or more, so as to call the corresponding parsing method to parse the associated consulting content info_0, and then obtain the first content feature of the text content corresponding to the consulting content. Exemplarily, the parsing method corresponding to the parsing identifier label_1 is to combine the text content based on the order of the moments, so according to the associated consulting content i The content generated by combining the text content corresponding to nfo_0 is "Purchase product A_1", and the corresponding first content feature is the purchase feature related to purchasing product A_1; similarly, when parsing the related consulting content info_1, it is necessary to match the corresponding parsing identifier label_2 for the case where the number of text contents is one, so as to call the corresponding parsing method to parse the related consulting content info_1, and then obtain the first content feature of the text content corresponding to the consulting content. By way of example, the parsing method corresponding to the parsing identifier label_2 is to parse the text content, so the text content "Product A_1" corresponding to the related consulting content info_1 is parsed, and it is determined that the corresponding first content feature is the product feature of product A_1.

[0060] In another possible implementation, Figure 3 for Figure 2 The schematic diagram of the specific implementation steps of step S102 in the embodiment shown is as follows: Figure 3 As shown, the specific implementation steps of step S102 include:

[0061] Step S1021: Describe the related consultation content in natural language according to the preset analysis information to obtain target analysis information.

[0062] Step S1022: The pre-trained large language model performs semantic analysis on the target parsed information to obtain a first content feature.

[0063] Exemplarily, for the text content "Product A_1" corresponding to the associated consulting content info_1, a natural language description is performed based on the preset parsing information, and the generated target parsing information text_1 is "If the user now says product A_1, what intention does the user want to express? Do not add any modifications, do not try to diverge intentions and fabricate replies, please use one sentence to accurately summarize the user's intentions", and then the target parsing information text_1 is input into the pre-trained large language model. The pre-trained large language model performs semantic analysis on the target parsing information text_1, and can output the corresponding first content feature. The first content feature is "The user's intention is to obtain the product features of product A_1, including the product introduction, user evaluation and purchase link of product A_1".

[0064] Similarly, for the text content "product A_1" and the text content "purchase" corresponding to the associated consulting content info_0, a natural language description is performed based on the preset parsing information, and the generated target parsing information text_2 is "The user has already said product A_1. If the user now says buy, what intention does the user want to express? Do not add any modifications, do not try to diverge intentions and fabricate replies, please use one sentence to accurately summarize the user's intentions." The target parsing information text_2 is then input into the pre-trained large language model. The pre-trained large language model performs semantic analysis on the target parsing information text_2, and can output the corresponding first content feature, which is "The user's intention is: to purchase product A_1."

[0065] Furthermore, in yet another possible implementation, Figure 4 for Figure 3 The schematic diagram of the specific implementation steps of step S1021 in the embodiment shown is as follows: Figure 4 As shown, the specific implementation steps of step S1021 include:

[0066] Step S10211: Identify the business characteristics of the related consulting content to obtain a business characteristic identifier.

[0067] Step S10212: Use preset parsing information to match the service feature identifier to determine feature parsing information.

[0068] Step S10213: Describe the related consultation content in natural language according to the feature analysis information to obtain target analysis information.

[0069] Exemplarily, the business features of different businesses are different. Under different business features, the corresponding parsing identifiers are matched, and then the corresponding natural language content is output, so as to be used for the preset large language model to generate the first content feature. Specifically, for example, Figure 5A schematic diagram of a process for determining a first content feature based on a service feature identifier provided in this application, such as Figure 5 As shown, the text content corresponding to the associated consulting content info_3 is text_3, and then the business feature identifier label_3 is obtained based on the business features of the business corresponding to the text content text_3; the text content corresponding to the associated consulting content info_4 is text_4, and then the business feature identifier label_4 is obtained based on the business features of the business corresponding to the text content text_4; the text content corresponding to the associated consulting content info_5 is text_5, and then the business feature identifier label_3 is obtained based on the business features of the business corresponding to the text content text_5; furthermore, the preset parsing information includes feature parsing information anal_1, feature parsing information anal_2, and feature parsing information anal_3. Based on the matching of the business feature identifier label_3 with the feature parsing information anal_2, based on the business feature identifier label_3 el_4 matches the feature resolution information anal_3, that is, the feature resolution information anal_2 is determined from the preset resolution information according to the business feature identifier label_3, and the feature resolution information anal_3 is determined from the preset resolution information according to the business feature identifier label_4; then, the related consulting content info_3 is parsed through the feature resolution information anal_2 to obtain the corresponding natural language content, that is, the target resolution information info_G_3, the related consulting content info_5 is parsed through the feature resolution information anal_2 to obtain the corresponding natural language content, that is, the target resolution information info_G_5, and the related consulting content info_4 is parsed through the feature resolution information anal_3 to obtain the corresponding natural language content, that is, the target resolution information info_G_4.

[0070] In the steps of this embodiment, the business characteristics of the business corresponding to the related consulting content are identified, and then the corresponding feature resolution information is matched from the preset resolution information. Furthermore, the related consulting content matching it is analyzed through the feature resolution information to obtain the target resolution information, and then the large language model pre-trained in the subsequent step performs semantic analysis on the target resolution information to generate the first content feature, thereby improving the accuracy of the generation of the first content feature.

[0071] Step S103: Use a preset knowledge base to match one or more first content features to obtain target reply information.

[0072] Exemplarily, the target reply information is used to indicate the reply content of the intelligent customer service to reply the target user through the target client; after obtaining the first content feature, the server matches one or more first content features from the preset knowledge base, and obtains the reply content that the intelligent customer service wants to reply to the target customer, that is, the target reply information. In a possible implementation, if the first content feature is a purchase feature related to the purchase of product A_1, then according to the purchase feature corresponding to product A_1, the content related to the purchase feature corresponding to product A_1 is matched from the preset knowledge base, such as the purchase link of product A_1, the precautions for the first use of product A_1, and the user evaluation of product A_1, and the reply content that the intelligent customer service wants to reply to the target customer is obtained, that is, the target reply information is obtained; if the first content feature is the product feature of product A_1, then according to the product feature corresponding to product A_1, the content related to the product feature corresponding to product A_1 is matched from the preset knowledge base, such as the product parameters of product A_1, the usage introduction of product A_1, the user evaluation of product A_1, and the purchase link of product A_1 are obtained, and the reply content that the intelligent customer service wants to reply to the target customer is obtained, that is, the target reply information is obtained.

[0073] Step S104, obtaining pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time.

[0074] Step S105: extract one or more second content features from the pre-consultation content.

[0075] Step S106: Calculate the similarity between each first content feature and each second content feature to obtain content similarity.

[0076] Step S107 , filtering out the predicted consultation content for the target user from the second content feature according to the content similarity.

[0077] Exemplarily, the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time; wherein the preset time interval is, for example, 5 minutes, 10 minutes, or 1 hour; the predicted consultation content is used to indicate the next consultation content that the target user predicts to consult by the intelligent customer service. Specifically, for example, the server determines the pre-consultation content within the corresponding time range based on the consultation time and the preset time interval of the target user, and extracts one or more second content features from the pre-consultation content through semantic analysis; then, by calculating the vector distance between each first content feature and each second content feature, the similarity between each first content feature and each second content feature is determined, that is, the content similarity is obtained; then, the second content features are screened according to the content similarity, and the predicted consultation content for the target user is generated based on the second content feature with the highest content similarity.

[0078] In the steps of this embodiment, the pre-consultation content within the corresponding preset time range is determined by the consultation time, and then the degree of correlation between the target consultation information and the pre-consultation information, that is, the content similarity, is determined by calculating the vector distance between each first content feature and each second content feature. Then, based on the second content feature with the highest content similarity, the next consultation content that the target user wants to consult is predicted by the intelligent customer service, that is, the predicted consultation content, so that it can be displayed on the target client of the target user, thereby improving the user experience.

[0079] Furthermore, in a possible implementation manner, the specific implementation steps of step S106 also include:

[0080] Step S1061, calculating the user similarity between the target user and other users.

[0081] Step S1062: Screen the second content features according to the user similarity to obtain at least two target second content features.

[0082] Step S1063 , calculating the similarity between each first content feature and each target second content feature to obtain content similarity.

[0083] Exemplarily, after receiving multiple pending consulting contents sent by the target client of the target user, the user registration information of the target user is obtained, and then the user similarity between the target user and other users is obtained by calculating the user registration information of the target user and the user registration information of other users, wherein the user registration information of other users is pre-stored information; then, the second content feature is screened based on the user similarity to obtain at least two target second content features, wherein the target second content feature is used to indicate the pre-consulting contents sent by other users with high similarity to the target user and the corresponding second content features. Specifically, for example, based on the user similarity and the preset user similarity threshold, the target other user is screened and determined from other users, and the user similarity between the target other user and the target user is greater than the preset user similarity threshold, and then based on the pre-consulting contents corresponding to the target other user, the target second content feature is determined from the second content features; further, by calculating the vector distance between each first content feature and each target second content feature, the similarity between each first content feature and each target second content feature is determined, and the content similarity is obtained. By screening the second content feature based on user similarity, the data processing efficiency is improved, so that the intelligent customer service can quickly feedback the target reply information and predict the consultation content to the target user, thereby improving the user experience.

[0084] Step S108, sending the target reply information and predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and predicted consultation content, and displays a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content.

[0085] Exemplarily, by sending target reply information and predicted consultation content to the target client of the target user, the target reply information and predicted consultation content can be displayed in the interactive area of ​​the target client, and a first interactive control for sending the predicted consultation content and / or a second interactive control for secondary editing of the predicted consultation content are further displayed; wherein the first interactive control enables the user to directly send the predicted consultation content by clicking the control to obtain the reply content corresponding to the predicted consultation content; the second interactive control enables the user to secondary edit the predicted consultation content by clicking the control, and then send the secondary edited predicted consultation content to obtain the corresponding reply content.

[0086] In the steps of this embodiment, on the basis of replying the target reply information to the user, by displaying the predicted consulting content to the target user and providing a "direct send" interaction component and / or a "secondary edit" interaction component corresponding to the predicted consulting content, it is avoided that the user needs to enter a large amount of text content when entering the next consulting content. For example, the user originally needed to enter 100 characters. After the intelligent customer service provides the predicted consulting content, the user only needs to modify 20 characters based on the "secondary edit" interaction component, or does not need to modify, and can directly click the "direct send" interaction component, thereby improving the user experience.

[0087] In this embodiment, multiple consultation contents to be processed are received from a target client of a target user, wherein the consultation contents carry consultation time; consultation contents belonging to the same dialogue cycle are associated and merged according to the consultation time to obtain associated consultation contents; feature analysis is performed on the associated consultation contents according to preset analysis information to obtain one or more first content features corresponding to the associated consultation contents; one or more first content features are matched using a preset knowledge base to obtain target reply information; previous consultation contents are obtained according to the consultation time, wherein the previous consultation contents are consultation contents sent by other users through the client within a preset time interval before the consultation time; one or more second content features are extracted from the previous consultation content; the similarity between each first content feature and each second content feature is calculated to obtain content similarity; the predicted consultation content of the target user is filtered out from the second content features according to the content similarity; the target reply information and the predicted consultation content are sent to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consultation content, as well as a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content. By associating and merging the consultation contents according to the consultation time of the target user, the associated consultation contents are obtained; then, in combination with the resolution identifier corresponding to the preset resolution information, the associated consultation contents are feature analyzed to obtain the first content feature; further, the first content feature is matched with the preset knowledge base, so that the target reply information of the intelligent customer service to the target user can be determined from the preset knowledge base; further, by extracting the second content feature from the pre-consultation content and calculating the similarity between the first content feature and the second content feature, it is possible to filter out the predicted consultation content of the target user from the second content feature, and at the same time, provide the target user with a "direct send" interaction component and / or a "secondary editing" interaction component for the second content feature; the solution of the embodiment of the present application not only solves the problem that the intelligent customer service cannot accurately reply to the user with relevant reply information, but also solves the problem that the corresponding predicted consultation information cannot be provided to the user, thereby solving the problem of poor user experience caused by the solution of the prior art.

[0088] Furthermore, the method provided by the embodiment of the present application also includes: performing sentence splitting on the predictive consultation content to obtain one or more predicted candidate words; sending so that the target client of the target user displays one or more predicted candidate words, and displays a third interactive control for the target user to select one or more predicted candidate words.

[0089] Exemplarily, one or more predicted candidate words are obtained by sentence splitting of the predicted consultation content; then, the one or more predicted candidate words are sent to the target client of the target user for display, and a third interactive control for selecting one or more predicted candidate words is provided to the target user, so that the target user can select one or more predicted candidate words through the third interactive control to generate the next consultation content for the target user, thereby avoiding the need to manually edit all the consultation content when the target user makes further consultations, thereby improving the user experience.

[0090] Figure 6 A flowchart of a method for processing consultation content based on intelligent customer service is provided for another embodiment of the present application, such as Figure 6 As shown, the consultation content processing method based on intelligent customer service provided in this embodiment is Figure 2 On the basis of the consultation content processing method based on intelligent customer service provided in the illustrated embodiment, step S101 is further refined, and the consultation content processing method based on intelligent customer service provided in this embodiment includes the following steps:

[0091] Step S200: receiving a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry consultation time.

[0092] Step S201, the consultation contents belonging to the same dialogue cycle are associated and merged according to the consultation time to obtain initial consultation contents.

[0093] Step S202, obtaining historical consultation content; the historical consultation content is the consultation content sent by the client of other users before the consultation time, and the historical consultation content carries the historical consultation time.

[0094] Step S203: extracting one or more reference consultation contents from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time.

[0095] Exemplarily, based on the consultation time, the consultation contents belonging to the same dialogue cycle are merged to obtain the initial consultation content; the server further obtains the consultation contents sent by the clients of other users before the consultation time, that is, the historical consultation contents, according to the consultation time, wherein the historical consultation contents carry the historical consultation time; and then, by calculating the time correlation between the consultation time and the historical consultation time, one or more reference consultation contents are extracted from the historical consultation contents. Specifically, for example, between the time periods (historical consultation time) T_1 and T_2, the service provider of the target client launched the activity activity_1, and the consultation contents sent by the user between the time periods T_1 and T_2 are strongly related to the activity activity_1; and then, based on the target user's consultation time being T_3, when the consultation time T_3 is between the time periods T_1 and T_2, the reference consultation contents are composed of percent_1 of the historical consultation contents his_text_1 related to the activity activity_1, percent_2 of the historical consultation contents his_text_1 related to the activity activity_ 1 irrelevant historical consultation content his_text_2, the sum of percent_1 and percent_2 is 1; when the consultation time T_3 is after time T_2, the difference in days day_1 between the consultation time T_3 and time T_2 is calculated, and then the proportion relationship between the historical consultation content his_text_1 and the historical consultation content his_text_2 is adjusted based on the difference in days day_1, that is, increase percent_2, decrease percent_1, and ensure that the sum of percent_1 and percent_2 is 1. For example, the adjustment formula is shown in formula (1).

[0096]

[0097] Figure 7 A schematic diagram of reference consultation content determined based on different consultation times provided for this application, such as Figure 7As shown, when the consultation time T_3 is between the time period T_1 and T_2, percent_1 is 80% and percent_2 is 20%, that is, the historical consultation content his_text_1 related to activity_1 accounts for 80% of the reference consultation content, and the historical consultation content his_text_2 not related to activity_1 accounts for 20% of the reference consultation content; when the consultation time T_3 is after time T_2, the difference day_1 between the consultation time T_3 and time T_2 is calculated. Based on formula (1), when the difference day_1 between the consultation time T_3 and time T_2 is 10 days, percent_1 is 70% and percent_2 is 30%; When the difference in days between T_3 and time T_2 is 70 days, percent_1 is 10% and percent_2 is 90%; when the difference in days between consultation time T_3 and time T_2 is 80 days, percent_1 is 0% and percent_2 is 100%; when the difference in days between consultation time T_3 and time T_2 is 90 days, percent_1 is 0% and percent_2 is 100%, that is, the historical consultation content his_text_1 related to activity activity_1 accounts for 0% of the reference consultation content, and the historical consultation content his_text_2 not related to activity activity_1 accounts for 100% of the reference consultation content.

[0098] In another possible implementation, Figure 8 for Figure 6 The schematic diagram of the specific implementation steps of step S203 in the embodiment shown is as follows: Figure 8 As shown, the specific implementation steps of step S203 include:

[0099] Step S2030, dividing the historical consultation time into intervals according to the preset start time and the preset end time to obtain a target time interval.

[0100] Step S2031, obtaining a target time identifier according to the relative time position relationship of the consultation time within the target time interval.

[0101] Exemplarily, the target time interval includes weekly time interval, monthly time interval, quarterly time interval and annual time interval. Taking the monthly time interval as an example, the first day of each month is used as the preset start time, and the last day of each month is used as the preset end time to obtain the corresponding monthly time interval. Based on the specific date corresponding to the consultation time, the position of the consultation time in the monthly time interval is obtained, that is, the target time identifier is obtained. For example, if the consultation time is the 8th of a certain month, the consultation time is located in the early ten days of the monthly time interval. If the consultation time is the 18th of a certain month, the consultation time is located in the middle ten days of the monthly time interval. If the consultation time is the 28th of a certain month, the consultation time is located in the late ten days of the monthly time interval.

[0102] Step S2032, dividing the historical consultation content according to the target time interval to obtain interval consultation content.

[0103] Step S2033, extracting reference consultation content from the interval consultation content according to the target time identifier.

[0104] Exemplarily, the historical consultation contents are rearranged and reorganized by the target time interval to obtain the interval consultation contents; and then, the reference consultation contents can be determined from the interval consultation contents according to the target time identifier. Specifically, for example, Fig. 9 A schematic diagram of determining reference consultation content based on interval consultation content provided by this application, such as Fig. 9 As shown, taking the monthly time interval as an example, the historical consulting content includes the consulting content from July to October, and then, based on the monthly time interval, the historical consulting content is divided into 4 interval consulting content, namely interval consulting content M_1, interval consulting content M_2, interval consulting content M_3 and interval consulting content M_4, and further divided into the first ten days, the middle ten days and the last ten days; further, based on the consultation time being in the first ten days of the monthly time interval, the consulting content corresponding to the target time mark of the first ten days is determined from the consulting content in the interval consulting content M_1, the interval consulting content M_2, the interval consulting content M_3 and the interval consulting content M_4, that is, the reference consulting content is obtained.

[0105] In the steps of this embodiment, based on the interval consultation content obtained by dividing and reorganizing the historical consultation content according to the target time interval, the corresponding target time identifier is obtained according to the position of the consultation time in the target time interval, and then the reference consultation content can be determined from the interval consultation content based on the target time identifier, thereby improving the accuracy of subsequent steps in determining related consultation content based on the reference consultation content.

[0106] Step S204, the initial consulting content is associated with one or more reference consulting contents to obtain associated consulting contents.

[0107] Exemplarily, after obtaining the reference consultation content, by comparing the initial consultation content with the reference consultation content, the reference consultation content with the highest similarity (correlation) is associated as the associated consultation content. Specifically, for example, the number of characters in the initial consultation content is 20, and the accuracy of the target user's true intention that it can express is low, and the number of characters in each reference consultation content is more than 100, and the accuracy of the user's true intention that it can express is high, then the vector distance between the initial consultation content and the reference consultation content is calculated respectively, for example, the vector distance is calculated by cosine similarity or Euclidean distance, and then the smaller the vector distance, the higher the similarity between the initial consultation content and the reference consultation content, and then the reference consultation content with the highest similarity is determined as the associated consultation content.

[0108] Fig.10 for Figure 6 The schematic diagram of the specific implementation steps of step S204 in the embodiment shown is as follows: Fig.10 As shown, the specific implementation steps of step S204 include:

[0109] Step S2041, dividing the consultation time into a first consultation time interval and a second consultation time interval according to the time sequence.

[0110] Step S2042: Based on the first consultation time interval, perform sentence component analysis on the initial consultation content to obtain at least one first keyword.

[0111] Step S2043: Based on the second consultation time interval, perform sentence component analysis on the initial consultation content to obtain at least one second keyword.

[0112] Exemplarily, based on the chronological order of the consultation contents sent by the target user, the consultation time is divided into a first consultation time interval and a second consultation time interval; based on the first consultation time interval, the content of the corresponding time interval is determined from the initial consultation content, and then a sentence component analysis is performed to obtain at least one first keyword; based on the second consultation time interval, the content of the corresponding time interval is determined from the initial consultation content, and then a sentence component analysis is performed to obtain at least one second keyword; specifically, for example, if the initial consultation content sent by the target user contains interference words such as adjectives and adverbs, the sentence structure of the initial consultation content is deconstructed and analyzed through sentence component analysis, and the interference words are deleted and filtered to obtain the first keyword and the second keyword in the corresponding time interval.

[0113] Step S2044: performing correlation processing on one or more reference consultation contents according to at least one first keyword and at least one second keyword to obtain correlated consultation contents.

[0114] Exemplarily, on the basis of obtaining at least one first keyword and at least one second keyword, the first keyword and the second keyword are combined based on the word part of speech, that is, a keyword phrase (or keyword short sentence) is obtained, and then the vector distance between the keyword phrase (or keyword short sentence) and each reference consulting content is calculated, and then based on the vector distance, the reference consulting content with the highest similarity is determined as the related consulting content.

[0115] Step S205: performing feature analysis on the associated consulting content according to preset analysis information to obtain one or more first content features corresponding to the associated consulting content.

[0116] Step S206: Use a preset knowledge base to match one or more first content features to obtain target reply information.

[0117] Step S207, obtaining pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time.

[0118] Step S208: extract one or more second content features from the pre-consultation content.

[0119] Step S209: Calculate the similarity between each first content feature and each second content feature to obtain content similarity.

[0120] Step S210 , filtering out predicted consultation content for the target user from the second content feature according to the content similarity.

[0121] Step S211, sending the target reply information and predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and predicted consultation content, and displays a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content.

[0122] In this embodiment, the implementation method of step S200, step S205-step S211 is the same as that of the present application. Figure 2 The implementation methods of step S100 and step S102 to step S108 in the illustrated embodiment are the same and will not be described in detail herein.

[0123] Fig.11 A schematic diagram of the structure of a consultation content processing device based on intelligent customer service provided in one embodiment of the present application is shown in FIG. Fig.11 As shown, the consultation content processing device 3 based on intelligent customer service provided in this embodiment includes:

[0124] The receiving module 30 is used to receive a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry consultation time;

[0125] The first processing module 31 is used to associate and merge the consultation contents belonging to the same dialogue cycle according to the consultation time to obtain the associated consultation contents; perform feature analysis on the associated consultation contents according to the preset analysis information to obtain one or more first content features corresponding to the associated consultation contents; and use the preset knowledge base to match the one or more first content features to obtain the target reply information;

[0126] The second processing module 32 is used to obtain the pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time; extract one or more second content features from the pre-consultation content; calculate the similarity between each first content feature and each second content feature to obtain content similarity; and filter out the predicted consultation content of the target user from the second content features according to the content similarity;

[0127] The output module 33 is used to send the target reply information and the predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consultation content, and displays the first interactive control for sending the predicted consultation content, and / or the second interactive control for secondary editing of the predicted consultation content.

[0128] In one possible design, when the first processing module 31 performs feature analysis on the associated consulting content according to preset analysis information to obtain one or more first content features corresponding to the associated consulting content, it is specifically used to: perform natural language description on the associated consulting content according to the preset analysis information to obtain target analysis information; and perform semantic analysis on the target analysis information using a pre-trained large language model to obtain the first content feature.

[0129] In one possible design, when the first processing module 31 performs a natural language description on the related consulting content according to the preset resolution information to obtain the target resolution information, it is specifically used to: identify the business characteristics of the related consulting content to obtain the business characteristic identifier; match the business characteristic identifier with the preset resolution information to determine the characteristic resolution information; and perform a natural language description on the related consulting content according to the characteristic resolution information to obtain the target resolution information.

[0130] In one possible design, when the first processing module 31 associates and merges consulting contents belonging to the same conversation cycle according to the consultation time to obtain associated consulting contents, it is specifically used to: associate and merge consulting contents belonging to the same conversation cycle according to the consultation time to obtain initial consulting contents; obtain historical consulting contents; wherein the historical consulting contents are consulting contents sent by the clients of other users before the consultation time, and the historical consulting contents carry the historical consultation time; based on the time correlation between the consultation time and the historical consultation time, extract one or more reference consulting contents from the historical consulting contents; and associate the initial consulting contents respectively through one or more reference consulting contents to obtain associated consulting contents.

[0131] In one possible design, when the first processing module 31 extracts one or more reference consultation contents from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time, it is specifically used to: divide the historical consultation time into intervals according to a preset start time and a preset end time to obtain a target time interval; obtain a target time identifier according to the relative time position relationship of the consultation time in the target time interval; divide the historical consultation contents into consultation contents according to the target time interval to obtain interval consultation contents; and extract reference consultation contents from the interval consultation contents according to the target time identifier.

[0132] In a possible design, when the first processing module 31 associates the initial consulting content with one or more reference consulting contents to obtain associated consulting contents, it is specifically used to: divide the consulting time into a first consulting time interval and a second consulting time interval according to the chronological order; based on the first consulting time interval, perform sentence component analysis on the initial consulting content to obtain at least one first keyword; based on the second consulting time interval, perform sentence component analysis on the initial consulting content to obtain at least one second keyword; based on at least one first keyword and at least one second keyword, perform association processing on one or more reference consulting contents to obtain associated consulting contents.

[0133] In a possible design, the target time intervals include weekly time intervals, monthly time intervals, quarterly time intervals, and annual time intervals.

[0134] In a possible design, when the second processing module 32 calculates the similarity between each first content feature and each second content feature to obtain the content similarity, it is specifically used to: calculate the user similarity between the target user and other users; screen each second content feature according to the user similarity to obtain at least two target second content features; calculate the similarity between each first content feature and each target second content feature to obtain the content similarity.

[0135] In a possible design, the consulting content processing device 3 based on intelligent customer service is also used to: perform sentence splitting on the predicted consulting content to obtain one or more predicted candidate words; sending one or more predicted candidate words to the target client of the target user, so that the target client of the target user displays one or more predicted candidate words, and displays a third interactive control for the target user to select one or more predicted candidate words.

[0136] The receiving module 30, the first processing module 31, the second processing module 32 and the output module 33 are connected in sequence. The consultation content processing device 3 based on intelligent customer service provided in this embodiment can execute the following steps: Figure 2-Figure 10 The technical solutions of any of the method embodiments shown have similar implementation principles and technical effects, which will not be described in detail here.

[0137] The consulting content processing device based on intelligent customer service provided in the embodiment of the present application can be used to execute the technical solution of the consulting content processing method based on intelligent customer service in the above-mentioned embodiment. Its implementation principle and technical effect are similar and will not be repeated here.

[0138] It should be noted that it should be understood that the division of the various modules of the above device is only a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the first processing module 31 can be a separately established processing element, or it can be integrated in a chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a processing element of the above device. The function of the first processing module 31. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together, or they can be implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in the processor element or instructions in the form of software.

[0139] Fig.12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Fig.12 As shown, the electronic device may include: a transceiver 121 , a processor 122 , and a memory 123 .

[0140] The processor 122 executes the computer execution instructions stored in the memory, so that the processor 122 executes the scheme in the above embodiment. The processor 122 can be a general-purpose processor, including a central processing unit CPU, a network processor (network processor, NP), etc.; it can also be a digital signal processor DSP, an application-specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0141] The memory 123 is connected to the processor 122 via a system bus and completes communication between them. The memory 123 is used to store computer program instructions.

[0142] The transceiver 121 may be used to obtain tasks to be executed and configuration information of the tasks to be executed.

[0143] The system bus can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The system bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The transceiver is used to realize the communication between the database access device and other computers (such as the target client, the read-write library, and the read-only library). The memory may include random access memory (RAM) and may also include non-volatile memory.

[0144] The electronic device provided in the embodiment of the present application may be the terminal device of the above embodiment.

[0145] An embodiment of the present application also provides a chip for running instructions, which is used to execute the technical solution of the consultation content processing method based on intelligent customer service in the above embodiment.

[0146] An embodiment of the present application also provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are executed on a computer, the computer executes the technical solution of the consultation content processing method based on intelligent customer service in the above embodiment.

[0147] An embodiment of the present application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When at least one processor executes the computer program, it can implement the technical solution of the consultation content processing method based on intelligent customer service in the above embodiment.

[0148] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0149] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to implement the solution of this embodiment.

[0150] In addition, each functional module in each embodiment of the present application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The above-mentioned module-composed unit can be implemented in the form of hardware or in the form of hardware plus software functional units.

[0151] The above-mentioned integrated module implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform some steps of the methods of various embodiments of the present application.

[0152] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor.

[0153] The memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk, etc.

[0154] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0155] The above storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The storage medium can be any available medium that can be accessed by a general or special purpose computer.

[0156] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic control unit or a main control device.

[0157] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for processing consultation content based on intelligent customer service, characterized in that: Applied to service equipment, including: receiving a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry consultation time; According to the consultation time, the consultation contents belonging to the same dialogue cycle are associated and merged to obtain associated consultation contents; Performing feature analysis on the related consulting content according to preset analysis information to obtain one or more first content features corresponding to the related consulting content; Using a preset knowledge base to match the one or more first content features to obtain target reply information; Acquire pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time; extracting one or more second content features from the pre-consultation content; Calculating the similarity between each of the first content features and each of the second content features to obtain content similarity; Filtering the predicted consultation content of the target user from the second content feature according to the content similarity; Send the target reply information and the predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consultation content, and displays a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content.

2. The method according to claim 1, characterized in that The performing feature analysis on the related consulting content according to the preset analysis information to obtain one or more first content features corresponding to the related consulting content includes: Describing the related consultation content in natural language according to the preset analysis information to obtain target analysis information; The pre-trained large language model performs semantic analysis on the target parsed information to obtain the first content feature.

3. The method according to claim 2, characterized in that The step of describing the related consultation content in natural language according to the preset parsing information to obtain target parsing information includes: Identifying the business characteristics of the related consulting content to obtain a business characteristic identifier; Using the preset parsing information to match the service feature identifier to determine feature parsing information; The related consultation content is described in natural language according to the feature analysis information to obtain the target analysis information.

4. The method according to claim 1, characterized in that: The step of associating and merging the consultation contents belonging to the same dialogue cycle according to the consultation time to obtain associated consultation contents includes: According to the consultation time, the consultation contents belonging to the same dialogue cycle are associated and merged to obtain initial consultation contents; Obtain historical consultation content; wherein the historical consultation content is the consultation content sent by the client of other users before the consultation time, and the historical consultation content carries the historical consultation time; Extracting one or more reference consultation contents from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time; The initial consulting content is associated with the one or more reference consulting contents to obtain the associated consulting content.

5. The method according to claim 4, characterized in that The extracting one or more reference consultation contents from the historical consultation contents according to the time correlation between the consultation time and the historical consultation time includes: Divide the historical consultation time into intervals according to a preset start time and a preset end time to obtain a target time interval; Obtaining a target time identifier according to the relative time position relationship of the consultation time within the target time interval; Dividing the historical consultation content according to the target time interval to obtain interval consultation content; The reference consultation content is extracted from the interval consultation content according to the target time identifier.

6. The method according to claim 4, characterized in that The step of associating the initial consulting content with the one or more reference consulting contents to obtain the associated consulting content includes: Dividing the consultation time into a first consultation time interval and a second consultation time interval according to the time sequence; Based on the first consultation time interval, performing sentence component analysis on the initial consultation content to obtain at least one first keyword; Based on the second consultation time interval, performing sentence component analysis on the initial consultation content to obtain at least one second keyword; According to the at least one first keyword and the at least one second keyword, the one or more reference consultation contents are respectively processed for relevance to obtain the relevant consultation contents.

7. The method according to claim 5, characterized in that The target time interval includes a weekly time interval, a monthly time interval, a quarterly time interval and an annual time interval.

8. The method according to claim 1, characterized in that The calculating the similarity between each of the first content features and each of the second content features to obtain content similarity includes: Calculating user similarity between the target user and the other users; Screening each of the second content features according to the user similarity to obtain at least two target second content features; The similarity between each of the first content features and each of the target second content features is calculated to obtain the content similarity.

9. The method according to claim 1, characterized in that: The method further comprises: Sentence-splitting the predicted consultation content to obtain one or more predicted candidate words; The one or more predicted candidate words are sent to a target client of the target user, so that the target client of the target user displays the one or more predicted candidate words and displays a third interactive control for the target user to select the one or more predicted candidate words.

10. A consultation content processing device based on intelligent customer service, characterized in that: include: A receiving module, configured to receive a plurality of consultation contents to be processed sent by a target client of a target user, wherein the consultation contents carry a consultation time; A first processing module is used to associate and merge consultation contents belonging to the same conversation cycle according to the consultation time to obtain associated consultation contents; Performing feature analysis on the related consulting content according to preset analysis information to obtain one or more first content features corresponding to the related consulting content; Using a preset knowledge base to match the one or more first content features to obtain target reply information; A second processing module is used to obtain pre-consultation content according to the consultation time, wherein the pre-consultation content is the consultation content sent by other users through the client within a preset time interval before the consultation time; extracting one or more second content features from the pre-consultation content; Calculating the similarity between each of the first content features and each of the second content features to obtain content similarity; and selecting the predicted consultation content of the target user from the second content features according to the content similarity; An output module is used to send the target reply information and the predicted consultation content to the target client of the target user, so that the target client of the target user displays the target reply information and the predicted consultation content, and displays a first interactive control for sending the predicted consultation content, and / or a second interactive control for secondary editing of the predicted consultation content.

11. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 9 when executed by a processor.

13. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 9 when being executed by a processor.

Images