Interaction method, device, server and storage medium based on content library
By introducing multi-dimensional space and Boston matrix into the content library, the problem of single level and type of interactive content library is solved, the accuracy and generalization capabilities of intelligent interaction are improved, and differentiated user services are provided.
Patent Information
- Application Number
- CN202011035044.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-09-27
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2040-09-27
AI Technical Summary
The existing interactive content library has a relatively single interactive content hierarchy and type, and cannot provide corresponding content expression methods for users of different types and levels, resulting in the inability to provide differentiated services, affecting the accuracy and generalization capabilities of intelligent interactions.
Through a content library based on multiple dimension spaces, including content hierarchy, content type, user type, content channel and content expression methods, the Boston matrix is used to determine the value of the target dimension space, call the matching search engine to obtain interactive content, and use the user to Feedback optimizes the content library and provides differentiated services.
It improves the accuracy and generalization capabilities of intelligent interactions, and can provide differentiated interactive content according to user types and scenarios to meet the needs of different users.
Smart Images

Figure CN112182350B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to the field of computer technology, and in particular, to an interactive method, device, server, and storage medium based on a content library. Background Art
[0002] With the rapid development of the Internet and artificial intelligence technology, intelligent interaction plays an increasingly important role in user experience and information acquisition. Among them, the interactive content library is a key component for realizing intelligent interaction. However, the existing interactive content library has problems such as the level and type of interactive content is relatively single, and it cannot provide corresponding content expression methods for different types of users, and cannot provide users with different types and levels of service capabilities. Summary of the invention
[0003] The present invention provides an interactive method, device, server and storage medium based on a content library, so as to provide differentiated services to users through a content library based on multiple dimensional spaces, thereby improving the accuracy and generalization ability of intelligent interaction.
[0004] In a first aspect, an embodiment of the present invention provides an interactive method based on a content library, which is applied to a server and includes:
[0005] In response to an interaction request input by a user on a front-end interaction interface, determining a target scenario type corresponding to the interaction request;
[0006] According to the Boston Matrix of the content library, determine the target values of multiple dimensional spaces corresponding to the target scenario type. The dimensional spaces include: content level, content type, user type, content channel, and content expression method;
[0007] The search engine matching the target content type is called to obtain interactive content matching the target content level from the content library, and the interactive content is processed accordingly according to the target content expression method and fed back to the user.
[0008] Optionally, before responding to an interaction request input by a user on a front-end interaction interface and determining a target scenario type corresponding to the interaction request, the method further includes:
[0009] Acquire multiple preset dimensional spaces, and construct an initial content library matching the multiple dimensional spaces;
[0010] Obtain content pairs that match the user's interactive behavior, and update the initial content library based on each content pair;
[0011] According to the user interaction logs, the initial content library is iteratively optimized to obtain the final content library.
[0012] Optionally, obtain content pairs that match the user interaction behavior, including:
[0013] In response to a target interaction request sent by a user on a front-end interaction interface, feeding back to the user at least one candidate interaction content corresponding to the target interaction request;
[0014] Determining target interaction content corresponding to the target interaction request according to the user's positive feedback information on the candidate interaction content;
[0015] An association relationship is established between the target interaction request and the target interaction content as a content pair.
[0016] Optionally, update the initial content library based on each content pair, including:
[0017] According to the target interaction request in the content pair, the initial content library is checked for duplicates;
[0018] If there is a standard interaction request similar to the target interaction request in the initial content library, the target interaction request is merged with the standard interaction request, and the target interaction content in the content pair is merged with the standard interaction content corresponding to the standard interaction request;
[0019] If there is no standard interaction request similar to the target interaction request in the initial content library, information corresponding to each dimensional space is generated according to the content pair and added to the initial content library.
[0020] Optionally, in descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content;
[0021] The target interaction content in the content pair is merged with the standard interaction content corresponding to the standard interaction request, including:
[0022] Obtaining the content hierarchy corresponding to the target interactive content and the standard interactive content;
[0023] If the priority of the content level of the target interactive content is higher than the priority of the content level of the standard interactive content, the standard interactive content is replaced with the target interactive content;
[0024] If the priority of the content hierarchy of the target interactive content is lower than or equal to the priority of the content hierarchy of the standard interactive content, the standard interactive content is kept unchanged.
[0025] Optionally, the initial content library is iteratively optimized based on the user interaction log to obtain the final content library, including:
[0026] Obtain user interaction logs at preset time intervals;
[0027] Calculate the user's satisfaction with the information in the content library based on the user feedback information in the current user interaction log;
[0028] According to the satisfaction level, the content library is optimized until the user's satisfaction with the information in the content library is the highest, and the final content library is obtained.
[0029] Optionally, the content library is optimized according to the satisfaction until the user's satisfaction with the information in the content library is the highest, and a final content library is obtained, including:
[0030] According to the satisfaction, the values of multiple dimensional spaces corresponding to each scene type of the content library are adjusted, and / or the content distribution of the content library is adjusted until the user's satisfaction with the information in the content library is the highest, thereby obtaining the final content library.
[0031] Optionally, the user's satisfaction with the information in the content library is calculated based on the user feedback information in the current user interaction log, including:
[0032] Count the number of positive feedback messages and the total number of interactions in the current user interaction log;
[0033] The ratio of the number of positive feedback messages to the total number of interactions is calculated as the user's satisfaction with the information in the content library.
[0034] Optionally, construct an initial content library that matches multiple dimensional spaces, including:
[0035] Establish the correlation between user type, content channel, content level and content expression mode, and initialize the values of user type, content channel, content level and content expression mode for various scenario types corresponding to the Boston matrix of the initial content library;
[0036] Among them, the content expression methods include formal dimension expression methods and content dimension expression methods. The formal dimension expression methods at least include: image / video, voice and text, and the content dimension expression methods include: specialization and de-specialization.
[0037] In a second aspect, an embodiment of the present invention further provides an interactive device based on a content library, applied to a server, comprising:
[0038] A first determination module, configured to determine a target scenario type corresponding to an interaction request input by a user on a front-end interaction interface;
[0039] The second determination module is used to determine target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional spaces include: content level, content type, user type, content channel and content expression method;
[0040] The interactive module is used to call the search engine that matches the target content type, obtain interactive content that matches the target content level from the content library, and process the interactive content according to the target content expression method and feed it back to the user.
[0041] Optionally, also include:
[0042] A construction module, used to obtain a plurality of preset dimensional spaces and construct an initial content library matching the plurality of dimensional spaces before determining a target scene type corresponding to the interaction request in response to an interaction request input by a user on a front-end interaction interface;
[0043] An updating module, used to obtain content pairs matching the user's interaction behavior and update the initial content library according to each content pair;
[0044] The optimization module is used to iteratively optimize the initial content library according to the user interaction log to obtain the final content library.
[0045] Optionally, update modules, including:
[0046] A feedback unit, configured to respond to a target interaction request sent by a user on a front-end interaction interface and to provide feedback to the user on at least one candidate interaction content corresponding to the target interaction request;
[0047] A determination unit, configured to determine target interaction content corresponding to the target interaction request according to the user's positive feedback information on the candidate interaction content;
[0048] The association unit is used to establish an association relationship between the target interaction request and the target interaction content as a content pair.
[0049] Optionally, update modules, including:
[0050] A duplicate checking unit, used for checking duplicates of the initial content library according to the target interaction request in the content pair;
[0051] a merging unit, configured to merge the target interaction request with the standard interaction request if there is a standard interaction request similar to the target interaction request in the initial content library, and merge the target interaction content in the content pair with the standard interaction content corresponding to the standard interaction request;
[0052] The adding unit is used to generate information corresponding to each dimensional space according to the content pair and add it to the initial content library if there is no standard interaction request similar to the target interaction request in the initial content library.
[0053] Optionally, in descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content;
[0054] A merging unit, used to: obtain the content hierarchy corresponding to the target interactive content and the standard interactive content;
[0055] If the priority of the content level of the target interactive content is higher than the priority of the content level of the standard interactive content, the standard interactive content is replaced with the target interactive content;
[0056] If the priority of the content hierarchy of the target interactive content is lower than or equal to the priority of the content hierarchy of the standard interactive content, the standard interactive content is kept unchanged.
[0057] In a third aspect, an embodiment of the present invention further provides a server, including:
[0058] one or more processors;
[0059] a storage device for storing one or more programs,
[0060] When one or more programs are executed by one or more processors, the one or more processors implement the content library-based interaction method provided by any embodiment of the present invention.
[0061] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the content library-based interaction method provided by any embodiment of the present invention.
[0062] According to the technical solution of the embodiment of the present invention, the server responds to the interaction request input by the user in the front-end interaction interface, determines the target scene type corresponding to the interaction request; determines the target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional space includes: content level, content type, user type, content channel and content expression method; calls the search engine matching the target content type, obtains the interactive content matching the target content level from the content library, and processes the interactive content accordingly according to the target content expression method and feeds it back to the user, which solves the problem of relatively single interactive content level and type existing in the existing interactive content library, and provides differentiated services to users through the content library based on multiple dimensional spaces, thereby improving the accuracy and generalization ability of intelligent interaction. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Figure 1 is a flow chart of an interactive method based on a content library in Embodiment 1 of the present invention;
[0064] Figure 2a is a flow chart of an interactive method based on a content library in Embodiment 2 of the present invention;
[0065] Figure 2bIt is a structural schematic diagram of a Boston matrix of an initial content library in the second embodiment of the present invention;
[0066] Figure 2c is a flow chart of obtaining a content pair in Embodiment 2 of the present invention;
[0067] Figure 2d This is a flowchart of updating an initial content library in Embodiment 2 of the present invention;
[0068] Figure 2e This is a flowchart of iterative optimization of an initial content library in Embodiment 2 of the present invention;
[0069] Figure 3 is a structural schematic diagram of an interactive device based on a content library in Embodiment 3 of the present invention;
[0070] Figure 4 It is a structural diagram of a server in Embodiment 4 of the present invention. DETAILED DESCRIPTION
[0071] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0072] Embodiment 1
[0073] Figure 1 This is a flow chart of an interactive method based on a content library in Embodiment 1 of the present invention. This embodiment is applicable to situations where human-computer interaction is performed based on a content library. The method can be performed by an interactive device based on a content library. The device can be implemented by hardware and / or software and can generally be integrated in a server that provides human-computer interaction services. Figure 1 As shown, the method includes:
[0074] Step 110: In response to an interaction request input by a user on a front-end interaction interface, a target scene type corresponding to the interaction request is determined.
[0075] In this embodiment, the user can be an ordinary customer or a business professional. The front-end interactive interface is a visual interface provided to the user. The user can communicate with the server through the front-end interactive interface to obtain interactive content fed back by the server. The interactive request can be a user's question about a certain business, for example, what to do if the credit card is lost, or a user's request to read information about a certain product.
[0076] In this embodiment, since ordinary customers and business professionals have different requirements for the interactive content of feedback, for example, ordinary users lack professional knowledge and the interactive content they obtain is usually for their own service. Therefore, they often expect the interactive content of feedback to be more understandable so as to facilitate their own understanding. The interactive content obtained by business professionals is often to serve ordinary customers or to increase their own professional accomplishments. Therefore, they often expect the interactive content of feedback to be more professional. Secondly, users have different requirements for the accuracy of interactive content in different scenarios. Therefore, various types of scenarios can be set according to the user type and the user's requirements for the interactive content of feedback, for example, scenarios for ordinary customers requiring easy-to-understand and accurate interactive content, scenarios for business professionals requiring professional and generalized interactive content, and so on.
[0077] Step 120: Determine target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library.
[0078] Among them, the dimensional space includes: content level, content type, user type, content channel and content expression method.
[0079] In this embodiment, in order to provide differentiated services to users and improve the accuracy and generalization of intelligent interaction, the content library is defined based on multiple dimensional spaces, including content levels, content types, user types, content channels, and content expression methods. Among them, the content level is used to describe the accuracy and detail of the interactive content in a hierarchical manner, the content type is used to classify the business type, which may include information, work orders, knowledge, product descriptions, marketing cases, business navigation, business processes, and other types, the user type is used to distinguish different customer groups, which may include ordinary customers and business professionals, the content channels may include: PC terminals, mobile terminals, STM terminals, etc., and the content expression method is used to distinguish different customer groups from having different expressions for the same content. For example, for ordinary customers and business professionals, the expression methods of the same interactive content are different.
[0080] Furthermore, in order to facilitate the rapid generation of interactive content corresponding to each scenario, the Boston Matrix is used to define the content library under each scenario from multiple dimensional spaces. For example, a scenario is defined in which the interactive content required by ordinary customers is easy to understand and accurate, the content level is set as direct answer or content segment, the content type is set as information, knowledge, transaction navigation, etc., the user type is set as ordinary customer, the content channel is set as mobile terminal, and the content expression method is set as de-specialization and easy to understand.
[0081] In this embodiment, after determining the target scene type corresponding to the user's interaction request, the target scene type can be matched among various scene types corresponding to the Boston matrix of the content library, thereby determining the target values of multiple dimensional spaces corresponding to the target scene type, so as to facilitate subsequent search for interactive content corresponding to the interaction request based on the target values of multiple dimensional spaces and corresponding processing of the interactive content.
[0082] Step 130: call a search engine that matches the target content type, obtain interactive content that matches the target content level from the content library, and process the interactive content accordingly according to the target content expression method and feed it back to the user.
[0083] In this embodiment, in order to improve the search accuracy of interactive content, dedicated search engines can be set for various content types, so that after determining the target values of multiple dimensional spaces corresponding to the target scene type, the search engine matching the target content type can be called to accurately search for interactive content matching the target content level from the content library, and then the interactive content is processed accordingly according to the target content expression method, for example, the interactive content is de-professionalized to meet the needs of users, and the processed interactive content is fed back to users to assist users in handling business or to answer questions for users.
[0084] In this embodiment, since different users have different understanding and reading habits for the same content information, the information is expressed differently for different customer groups. The expression method can be defined in two dimensions, one is formal, which is commonly in the form of text, language, graphics, images, sounds, etc., and the other is content, such as professional and non-professional. For example, in the financial field, when ordinary customers consult questions, they hope that the feedback content can block difficult-to-understand professional language as much as possible, while for business professionals, it is more concise and efficient to use professional descriptions. In the process of intelligent interaction, the content expression method of the interactive content returned for the same question can be matched based on user type and content channel to return interactive content that is more suitable for users.
[0085] According to the technical solution of the embodiment of the present invention, the server responds to the interaction request input by the user in the front-end interaction interface, determines the target scene type corresponding to the interaction request; determines the target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional space includes: content level, content type, user type, content channel and content expression method; calls the search engine matching the target content type, obtains the interactive content matching the target content level from the content library, and processes the interactive content accordingly according to the target content expression method and feeds it back to the user, which solves the problem of relatively single interactive content level and type existing in the existing interactive content library, and provides differentiated services to users through the content library based on multiple dimensional spaces, thereby improving the accuracy and generalization ability of intelligent interaction.
[0086] Embodiment 2
[0087] Figure 2a This is a flow chart of an interactive method based on a content library in Embodiment 2 of the present invention. This embodiment is further refined on the basis of the above embodiment. Figure 2a An interactive method based on a content library provided in the second embodiment of the present application is described, comprising the following steps:
[0088] Step 210: Acquire multiple preset dimensional spaces, and construct an initial content library matching the multiple dimensional spaces.
[0089] In this embodiment, in order to realize a content library with multiple levels, multiple types, and diversified content expressions and to improve the service quality and efficiency of the content library, five dimensional spaces of content level, content type, user type, content channel, and content expression are pre-set, and the initial content library is defined and constructed based on these five dimensional spaces.
[0090] Optionally, constructing an initial content library that matches multiple dimensional spaces may include: establishing a correlation between user types, content channels, content levels and content expressions, and initializing the values of user types, content channels, content levels and content expressions for various scenario types corresponding to the Boston matrix of the initial content library; wherein the content expressions include formal dimension expressions and content dimension expressions, the formal dimension expressions include at least: images / videos, voices and texts, and the content dimension expressions include: specialization and de-specialization.
[0091] In this embodiment, the interrelationship between user type, content channel, content level and content expression can be established according to the requirements of different user types for interactive content. For example, business professionals usually use PCs for intelligent interaction, and the content expression is more expected to be in text form and more professional. In scenarios with high accuracy requirements, the interactive content needs to be a question answer or a content segment. Then, the Boston Matrix method is used to determine the various scene types corresponding to the Boston Matrix of the initial content library from five dimensions, as well as the initial values of each dimensional space under each scene type, such as Figure 2b shown.
[0092] In this embodiment, the Boston Matrix of the initial content library describes the initial content library from two dimensions: the readability of the interactive content and the generalization ability of the interactive content. Figure 2b As shown, the first quadrant can correspond to the scenario where ordinary customers use intelligent interaction for business consultation or business processing, and the interactive content returned in this scenario is required to be easy to understand and relatively accurate, for example, it can accurately navigate to the business processing function, or accurately answer questions for users. Correspondingly, the content level corresponding to the first quadrant can be question answers and content segments, the content type can be any type, the user type is an ordinary customer, the content channel is a mobile terminal, and in terms of content expression, the expected order of content form is image / video, voice, and text form, and the content is expected to be more professional and easy to understand.
[0093] The second quadrant can correspond to the scenario where ordinary customers use intelligent interaction to read information or learn about product-related information. The interactive content returned in this scenario is required to be relatively complete and easy to understand. For example, users want to read consultation, knowledge information, etc. Correspondingly, the content level corresponding to the second quadrant can be documents and related content, the content type can be any type, the user type is ordinary customers, the content channel is mobile terminals, and in terms of content expression, the expected order of content form is image / video, voice, and text, and the content is expected to be more professional and easy to understand.
[0094] The third quadrant can correspond to scenarios where business professionals use intelligent interaction to read and modify information, and the interactive content returned in this scenario is required to be relatively complete and professional, for example, business personnel edit professional knowledge, answer customer questions, etc. Correspondingly, the content level corresponding to the third quadrant can be documents and related content, the content type can be any type, the user type is business professionals, the content channel is PC, and the content expression method is expected to be in text form and the content is more professional.
[0095] The fourth quadrant can correspond to the scenario where business professionals use intelligent interaction to quickly answer questions for users or help users handle business. The interactive content returned in this scenario is required to be relatively professional and accurate, for example, the scenario where the seat staff answers customer questions. Accordingly, the content level corresponding to the fourth quadrant can be question answers and content segments, the content type can be any type, the user type is business professionals, the content channel is PC, and the content expression method is expected to be in text form and the content is more professional.
[0096] In this embodiment, a front-end interactive interface is provided to users on a PC, mobile terminal or STM terminal, and intelligent interactive services are performed by calling a content library on a server. By defining a content library based on multiple dimensional spaces, different interactive scenarios can be met, thereby improving the scope and capabilities of intelligent interactive services.
[0097] Step 220: Obtain content pairs that match the user interaction behavior, and update the initial content library according to each content pair.
[0098] In this embodiment, after constructing an initial content library matching multiple dimensional spaces, the initial content library may be supplemented and updated based on content pairs generated by user interaction behaviors to obtain a more complete content library.
[0099] Optionally, obtaining a content pair that matches the user interaction behavior may include: in response to a target interaction request sent by the user on the front-end interaction interface, feeding back to the user at least one alternative interaction content corresponding to the target interaction request; determining the target interaction content corresponding to the target interaction request based on the user's positive feedback information on the alternative interaction content; and establishing an association relationship between the target interaction request and the target interaction content as a content pair.
[0100] In this embodiment, Figure 2c As shown, when the user sends a target interaction request on the front-end interaction interface, for example, the user enters a search term or a question sentence on the front-end interaction interface, or initiates a customer service call on the front-end interaction interface, at least one alternative interaction content corresponding to the target interaction request is obtained and fed back to the user through the front-end interaction interface. The user's feedback information on each alternative interaction content is obtained, for example, the number of stars lit by the user for the alternative interaction content, or the user likes or dislikes the alternative interaction information, or scores the alternative interaction content, or describes the alternative interaction content in text to express approval or complaints, and counts the positive feedback information. Then, the alternative interaction content corresponding to the positive feedback information is used as the target interaction content corresponding to the target interaction request, and the target interaction request is associated with the target interaction content to obtain a content pair that matches the user's interaction behavior.
[0101] For example, assuming that the user inputs "I can't find my credit card", when the intelligent question and answer cannot directly give an answer to the question, a list of search results is returned by searching for credit card-related documents, and each search result displays the optimal segment information; if the user takes an affirmative action on a certain result (such as the credit card loss report process as the optimal segment), a matching relationship can be established between "I can't find my credit card" and "Credit card loss report process as the optimal segment"; if the user clicks on a certain document (such as a credit card operation manual) and takes an affirmative action, a matching relationship can be established between "I can't find my credit card" and "Credit Card Operation Manual" to generate a content pair.
[0102] Optionally, updating the initial content library according to each content pair may include: checking the initial content library for duplicates according to the target interaction request in the content pair; if there is a standard interaction request similar to the target interaction request in the initial content library, merging the target interaction request with the standard interaction request, and merging the target interaction content in the content pair with the standard interaction content corresponding to the standard interaction request; if there is no standard interaction request similar to the target interaction request in the initial content library, generating information corresponding to each dimensional space according to the content pair, and adding it to the initial content library.
[0103] In this embodiment, after obtaining content pairs, that is, matching pairs of questions and answers at different levels, a duplicate check is performed in the initial content library according to the target interaction request in the content pair. If no standard interaction request similar to the target interaction request is found, the content pair is directly added to the initial content library, and information such as the content level, content type, user type, content channel, and content expression method generated according to the content pair is also added to the initial content library; if there is a standard interaction request similar to the target interaction request in the initial content library, the target interaction request and the target interaction content are merged with the standard interaction request and the standard interaction content, respectively, such as Figure 2d shown.
[0104] Optionally, in descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content;
[0105] Merging the target interactive content in the content pair with the standard interactive content corresponding to the standard interactive request may include: obtaining the content hierarchy corresponding to the target interactive content and the standard interactive content; if the priority of the content hierarchy of the target interactive content is higher than the priority of the content hierarchy of the standard interactive content, replacing the standard interactive content with the target interactive content; if the priority of the content hierarchy of the target interactive content is lower than or equal to the priority of the content hierarchy of the standard interactive content, keeping the standard interactive content unchanged.
[0106] In this embodiment, the content level is used to describe the accuracy and detail of the interactive content in a hierarchical manner, so the priority of the answer to the question, the content segment, the document, and the related content is from high to low. When the target interactive content and the standard interactive content are merged, the priority of the content level corresponding to the target interactive content and the standard interactive content can be determined first. If the priority of the content level of the target interactive content is higher, that is, the accuracy and detail of the target interactive content are better, the standard interactive content is replaced by the target interactive content. Otherwise, that is, the accuracy and detail of the standard interactive content are better, the standard interactive content is kept unchanged.
[0107] Exemplarily, for the content pair "I can't find my credit card" and "Credit card loss reporting process search optimal segment", "I can't find my credit card" is checked for duplication in the initial content library. If no similar duplicate questions such as "Credit card lost" are found, "Credit card loss reporting process search optimal segment" is directly added to the initial content library. If similar duplicate questions such as "Credit card lost" are found, the question "I can't find my credit card" is used as an extended question and merged with the current question "How to deal with the loss of a credit card". The merging of interactive content should be carried out according to the priority of the content hierarchy. If the standard interactive content is the answer to the question of credit card loss, it will not be merged. If the standard interactive content is a credit card operation manual, "Credit card loss reporting process search optimal segment" will replace the standard interactive content in the initial content library.
[0108] In this embodiment, by generating a content library of multiple types, multiple levels, multiple expressions, and dynamically adjustable content, the content library has the characteristics of wide application range and providing differentiated services. By updating the content library, the content in the content library is made real-time.
[0109] Step 230: Iteratively optimize the initial content library according to the user interaction log to obtain a final content library.
[0110] In this embodiment, after the initial content library has been running for a period of time, the content therein increases, and batch iterative optimization is required to ensure that the information in the initial content library is optimized. The iterative optimization of the initial content library is also a process of gradually adjusting the four quadrant definitions corresponding to the Boston Matrix of the initial content library.
[0111] Optionally, the initial content library is iteratively optimized based on the user interaction log to obtain the final content library, which may include: obtaining the user interaction log at preset time intervals; calculating the user's satisfaction with the information in the content library based on the user feedback information in the current user interaction log; and optimizing the content library based on the satisfaction until the user's satisfaction with the information in the content library is the highest, thereby obtaining the final content library.
[0112] In this embodiment, the user interaction log refers to the interactive flow data information generated when the user uses the search and intelligent question-answering interactive functions, such as input questions, search terms, content channels, user types, etc. Figure 2e As shown, the user interaction log can be obtained once a day and mapped to the five-dimensional space of the content library. Then, the user feedback information on the interaction content is extracted from the user interaction log, and the user satisfaction with the interaction content is calculated. Finally, the content library is optimized according to the five dimensions of each quadrant corresponding to the Boston matrix of the content library and the user satisfaction, so as to maximize the user satisfaction in the content library and obtain the final content library.
[0113] Optionally, calculating the user's satisfaction with the information in the content library based on the user feedback information in the current user interaction log can include: counting the number of positive feedback messages and the total number of interactions in the current user interaction log; and calculating the ratio of the number of positive feedback messages to the total number of interactions as the user's satisfaction with the information in the content library.
[0114] In this embodiment, the ratio of positive feedback information that meets the preset conditions to the total number of interactions can be used as the user's satisfaction with the information in the content library, where the preset conditions can be that the number of stars lit by the user is greater than or equal to 80% of the total number of stars, the user likes the feedback interactive content, or the user's score for the interactive content is greater than 70% of the total score, etc. The ratio of the number of positive feedback information to the number of negative feedback information can also be used as the user's satisfaction with the information in the content library, etc.
[0115] Optionally, the content library is optimized based on satisfaction until the user is most satisfied with the information in the content library, thereby obtaining the final content library. This may include: adjusting the values of multiple dimensional spaces corresponding to each scene type of the content library, and / or adjusting the content distribution of the content library based on satisfaction, until the user is most satisfied with the information in the content library, thereby obtaining the final content library.
[0116] In this embodiment, Figure 2e As shown, the values of multiple dimensional spaces corresponding to each scenario type of the content library can be adjusted according to satisfaction. For example, if the user satisfaction in the second quadrant is low, it may be that the answers given to users are too generalized, with many inaccurate results, and the priority of the content level needs to be improved; if the user in the first quadrant is dissatisfied, it may be that the response rate of the answer to the question is low, and it is necessary to consider expanding the content library. Adjusting the content distribution of the content library can be adjusting the service mode according to the user satisfaction. For example, for search and question and answer, the service mode for the user with higher user satisfaction can be set to that one. It can also be the adjustment of the matching threshold of the search. If the threshold is too high, many questions cannot be answered, and if the threshold is too low, wrong search results will be found.
[0117] Step 240: In response to an interaction request input by the user in the front-end interaction interface, determine a target scene type corresponding to the interaction request.
[0118] Step 250: Determine target values of multiple dimensional spaces corresponding to the target scene type according to the Boston Matrix of the content library, where the dimensional spaces include: content level, content type, user type, content channel, and content expression method.
[0119] Step 260: call a search engine that matches the target content type, obtain interactive content that matches the target content level from the content library, and process the interactive content accordingly according to the target content expression method and feed it back to the user.
[0120] According to the technical solution of the embodiment of the present invention, the server responds to the interaction request input by the user in the front-end interaction interface, determines the target scene type corresponding to the interaction request; determines the target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional space includes: content level, content type, user type, content channel and content expression method; calls the search engine matching the target content type, obtains the interactive content matching the target content level from the content library, and processes the interactive content accordingly according to the target content expression method and feeds it back to the user, which solves the problem of relatively single interactive content level and type existing in the existing interactive content library, and provides differentiated services to users through the content library based on multiple dimensional spaces, thereby improving the accuracy and generalization ability of intelligent interaction.
[0121] Embodiment 3
[0122] Figure 3 is a schematic diagram of the structure of an interactive device based on a content library in Embodiment 3 of the present invention. This embodiment is applicable to situations where human-computer interaction is performed based on a content library. The device can be implemented by hardware and / or software and can generally be integrated in a server that provides human-computer interaction services. Figure 3 As shown, the device is applied to a server and includes:
[0123] A first determination module 310 is used to determine a target scenario type corresponding to an interaction request in response to an interaction request input by a user on a front-end interaction interface;
[0124] A second determination module 320 is used to determine target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional spaces include: content level, content type, user type, content channel and content expression method;
[0125] The interactive module 330 is used to call a search engine that matches the target content type, obtain interactive content that matches the target content level from the content library, and process the interactive content according to the target content expression mode and feed it back to the user.
[0126] According to the technical solution of the embodiment of the present invention, the server responds to the interaction request input by the user in the front-end interactive interface, determines the target scene type corresponding to the interaction request; determines the target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, and the dimensional space includes: content level, content type, user type, content channel and content expression method; calls the search engine matching the target content type, the content library obtains the interactive content matching the target content level, and processes the interactive content accordingly according to the target content expression method and feeds it back to the user, which solves the problem of relatively single interactive content level and type existing in the existing interactive content library, and provides differentiated services to users through the content library based on multiple dimensional spaces, thereby improving the accuracy and generalization ability of intelligent interaction.
[0127] Optionally, also include:
[0128] A construction module, used to obtain a plurality of preset dimensional spaces and construct an initial content library matching the plurality of dimensional spaces before determining a target scene type corresponding to the interaction request in response to an interaction request input by a user on a front-end interaction interface;
[0129] An updating module, used to obtain content pairs matching the user's interaction behavior and update the initial content library according to each content pair;
[0130] The optimization module is used to iteratively optimize the initial content library according to the user interaction log to obtain the final content library.
[0131] Optionally, update modules, including:
[0132] A feedback unit, configured to respond to a target interaction request sent by a user on a front-end interaction interface and to provide feedback to the user on at least one candidate interaction content corresponding to the target interaction request;
[0133] A determination unit, configured to determine target interaction content corresponding to the target interaction request according to the user's positive feedback information on the candidate interaction content;
[0134] The association unit is used to establish an association relationship between the target interaction request and the target interaction content as a content pair.
[0135] Optionally, update modules, including:
[0136] A duplicate checking unit, used for checking duplicates of the initial content library according to the target interaction request in the content pair;
[0137] a merging unit, configured to merge the target interaction request with the standard interaction request if there is a standard interaction request similar to the target interaction request in the initial content library, and merge the target interaction content in the content pair with the standard interaction content corresponding to the standard interaction request;
[0138] The adding unit is used to generate information corresponding to each dimensional space according to the content pair and add it to the initial content library if there is no standard interaction request similar to the target interaction request in the initial content library.
[0139] Optionally, in descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content;
[0140] A merging unit, used to: obtain the content hierarchy corresponding to the target interactive content and the standard interactive content;
[0141] If the priority of the content level of the target interactive content is higher than the priority of the content level of the standard interactive content, the standard interactive content is replaced with the target interactive content;
[0142] If the priority of the content hierarchy of the target interactive content is lower than or equal to the priority of the content hierarchy of the standard interactive content, the standard interactive content is kept unchanged.
[0143] Optional, optimization modules, including:
[0144] A log acquisition unit, used to acquire user interaction logs at preset time intervals;
[0145] A calculation unit, used to calculate the user's satisfaction with the information in the content library according to the user feedback information in the current user interaction log;
[0146] The optimization processing unit is used to optimize the content library according to the satisfaction level until the user's satisfaction with the information in the content library is the highest, thereby obtaining a final content library.
[0147] Optionally, the optimization processing unit is used to: adjust the values of multiple dimensional spaces corresponding to each scene type of the content library according to satisfaction, and / or adjust the content distribution of the content library until the user's satisfaction with the information in the content library is highest, thereby obtaining the final content library.
[0148] Optionally, the calculation unit is used to: count the number of positive feedback messages and the total number of interactions in the current user interaction log; and calculate the ratio of the number of positive feedback messages to the total number of interactions as the user's satisfaction with the information in the content library.
[0149] Optionally, a construction module is used to: establish a correlation between user type, content channel, content level and content expression method, and initialize the values of user type, content channel, content level and content expression method for various scenario types corresponding to the Boston matrix of the initial content library;
[0150] Among them, the content expression methods include formal dimension expression methods and content dimension expression methods. The formal dimension expression methods at least include: image / video, voice and text, and the content dimension expression methods include: specialization and de-specialization.
[0151] The content library-based interaction device provided in the embodiment of the present invention can execute the content library-based interaction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0152] Embodiment 4
[0153] Figure 4 A schematic diagram of the structure of a server disclosed in an embodiment of the present invention. Figure 4 A block diagram of an exemplary device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 4 The device 12 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0154] like Figure 4 As shown, device 12 is in the form of a general purpose computing device. Components of device 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, and bus 18 that connects various system components (including system memory 28 and processing unit 16).
[0155] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.
[0156] Device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by device 12, including volatile and non-volatile media, removable and non-removable media.
[0157] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.
[0158] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in the memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 42 generally perform the functions and / or methods of the embodiments described herein.
[0159] The device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may communicate with one or more devices that enable a user to interact with the device 12, and / or may communicate with any device that enables the device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the device 12 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the device 12 via a bus 18. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the device 12, including, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0160] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the content library-based interaction method provided in the embodiment of the present invention.
[0161] That is, to implement an interactive method based on a content library, applied to a server, including:
[0162] In response to an interaction request input by a user on a front-end interaction interface, determining a target scenario type corresponding to the interaction request;
[0163] According to the Boston Matrix of the content library, determine the target values of multiple dimensional spaces corresponding to the target scenario type. The dimensional spaces include: content level, content type, user type, content channel, and content expression method;
[0164] The search engine matching the target content type is called to obtain interactive content matching the target content level from the content library, and the interactive content is processed accordingly according to the target content expression method and fed back to the user.
[0165] Embodiment 5
[0166] Embodiment 5 of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a computer processor, it is used to execute an interactive method based on a content library, which is applied to a server and includes:
[0167] In response to an interaction request input by a user in a front-end interaction interface, a target scenario type corresponding to the interaction request is determined; based on the Boston Matrix of the content library, target values of multiple dimensional spaces corresponding to the target scenario type are determined, and the dimensional spaces include: content level, content type, user type, content channel, and content expression method; a search engine matching the target content type is called to obtain interactive content matching the target content level from the content library, and according to the target content expression method, the interactive content is processed accordingly and fed back to the user.
[0168] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.
[0169] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0170] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0171] Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0172] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. An interactive method based on a content library, characterized in that: Applicable to servers, including: Acquire multiple preset dimensional spaces, and construct an initial content library matching the multiple dimensional spaces; Acquire content pairs matching the user interaction behavior, and update the initial content library according to each of the content pairs; Iteratively optimizing the initial content library according to the user interaction log to obtain a final content library; In response to an interaction request input by a user on a front-end interaction interface, determining a target scene type corresponding to the interaction request; According to the Boston Matrix of the content library, target values of multiple dimensional spaces corresponding to the target scene type are determined, wherein the dimensional spaces include: content level, content type, user type, content channel, and content expression method; Invoke a search engine that matches the target content type, obtain interactive content that matches the target content level from a content library, and process the interactive content accordingly according to the target content expression method and feed it back to the user; The step of constructing an initial content library matching multiple dimensional spaces includes: Establishing the mutual correlation between user type, content channel, content level and content expression mode, and respectively initializing the values of user type, content channel, content level and content expression mode for various scene types corresponding to the Boston matrix of the initial content library; wherein the content expression mode includes a form dimension expression mode and a content dimension expression mode; the form dimension expression mode includes at least: image / video, voice and text, and the content dimension expression mode includes: specialization and de-specialization; The step of acquiring a content pair matching the user interaction behavior includes: In response to a target interaction request sent by a user on a front-end interaction interface, feeding back to the user at least one candidate interaction content corresponding to the target interaction request; Determining target interaction content corresponding to the target interaction request according to the positive feedback information of the user on the candidate interaction content; An association relationship is established between the target interaction request and the target interaction content as a content pair.
2. The method according to claim 1, characterized in that: Updating the initial content library according to each of the content pairs includes: Performing a duplicate checking process on the initial content library according to the target interaction request in the content pair; If a standard interaction request similar to the target interaction request exists in the initial content library, the target interaction request is merged with the standard interaction request, and the target interaction content in the content pair is merged with the standard interaction content corresponding to the standard interaction request; If there is no standard interaction request similar to the target interaction request in the initial content library, information corresponding to each of the dimensional spaces is generated according to the content pair and added to the initial content library.
3. The method according to claim 2, characterized in that In descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content; The target interaction content in the content pair is merged with the standard interaction content corresponding to the standard interaction request, including: Acquire the target interactive content and the content level corresponding to the standard interactive content; If the priority of the content level of the target interactive content is higher than the priority of the content level of the standard interactive content, replacing the standard interactive content with the target interactive content; If the priority of the content level of the target interactive content is lower than or equal to the priority of the content level of the standard interactive content, the standard interactive content is kept unchanged.
4. The method according to claim 1, characterized in that: According to the user interaction log, the initial content library is iteratively optimized to obtain a final content library, including: Obtain user interaction logs at preset time intervals; Calculating user satisfaction with the information in the content library based on user feedback information in the current user interaction log; According to the satisfaction level, the content library is optimized until the user's satisfaction with the information in the content library is the highest, thereby obtaining a final content library.
5. The method according to claim 4, characterized in that According to the satisfaction, the content library is optimized until the user's satisfaction with the information in the content library is the highest, and a final content library is obtained, including: According to the satisfaction, the values of multiple dimensional spaces corresponding to each scene type of the content library are adjusted, and / or the content distribution of the content library is adjusted until the user is most satisfied with the information in the content library, thereby obtaining a final content library.
6. The method according to claim 4, characterized in that Calculating the user's satisfaction with the information in the content library according to the user feedback information in the current user interaction log, including: Counting the number of positive feedback messages and the total number of interactions in the current user interaction log; The ratio of the number of positive feedback messages to the total number of interactions is calculated as the user's satisfaction with the information in the content library.
7. The method according to claim 1, characterized in that Build an initial content library that matches multiple dimensional spaces, including: Establish the correlation between user type, content channel, content level and content expression mode, and initialize the values of user type, content channel, content level and content expression mode for various scenario types corresponding to the Boston matrix of the initial content library; The content expression method includes a form dimension expression method and a content dimension expression method. The form dimension expression method at least includes: image / video, voice and text. The content dimension expression method includes: specialization and de-specialization.
8. An interactive device based on a content library, characterized in that: Applicable to servers, including: A construction module, configured to obtain a plurality of preset dimensional spaces and construct an initial content library matching the plurality of dimensional spaces before determining a target scene type corresponding to the interaction request in response to an interaction request input by a user on a front-end interaction interface; An updating module, configured to obtain content pairs matching the user interaction behaviors, and update the initial content library according to each of the content pairs; An optimization module, configured to iteratively optimize the initial content library according to the user interaction log to obtain a final content library; A first determination module, configured to determine a target scenario type corresponding to an interaction request input by a user on a front-end interaction interface; A second determination module is used to determine target values of multiple dimensional spaces corresponding to the target scene type according to the Boston matrix of the content library, wherein the dimensional spaces include: content level, content type, user type, content channel, and content expression method; An interactive module, used to call a search engine that matches the target content type, obtain interactive content that matches the target content level from a content library, and process the interactive content according to the target content expression mode and feed it back to the user; The construction module is further used to establish the mutual correlation between user type, content channel, content level and content expression mode, and initialize the values of user type, content channel, content level and content expression mode respectively for various scene types corresponding to the Boston matrix of the initial content library; wherein the content expression mode includes a form dimension expression mode and a content dimension expression mode; the form dimension expression mode at least includes: image / video, voice and text, and the content dimension expression mode includes: specialization and de-specialization; Wherein, the update module includes: A feedback unit, configured to respond to a target interaction request sent by a user on a front-end interaction interface and to provide feedback to the user on at least one candidate interaction content corresponding to the target interaction request; A determination unit, configured to determine a target interaction content corresponding to the target interaction request according to the positive feedback information of the user on the candidate interaction content; The associating unit is used to establish an associative relationship between the target interaction request and the target interaction content as a content pair.
9. The device according to claim 8, characterized in that Update modules, including: A duplicate checking unit, configured to perform a duplicate checking process on the initial content library according to the target interaction request in the content pair; a merging unit, configured to merge the target interaction request with the standard interaction request if there is a standard interaction request similar to the target interaction request in the initial content library, and merge the target interaction content in the content pair with the standard interaction content corresponding to the standard interaction request; An adding unit is used to generate information corresponding to each of the dimensional spaces according to the content pair and add the information to the initial content library if there is no standard interaction request similar to the target interaction request in the initial content library.
10. The device according to claim 9, characterized in that In descending order of priority, the content hierarchy includes question answers, content segments, documents, and related content; A merging unit, used to: obtain the content level corresponding to the target interactive content and the standard interactive content; If the priority of the content level of the target interactive content is higher than the priority of the content level of the standard interactive content, replacing the standard interactive content with the target interactive content; If the priority of the content level of the target interactive content is lower than or equal to the priority of the content level of the standard interactive content, the standard interactive content is kept unchanged.
11. A server, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the content library-based interaction method as described in any one of claims 1 to 7.
12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the content library-based interaction method as described in any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Man-machine interaction method and device, server and storage medium
CN109961780A
