Question and answer processing method and system, electronic equipment and storage medium
By building basic data service components and a large language model for business scenarios, the problems of large data volume and high cost in question-and-answer processing were solved, enabling efficient and accurate user answers and improving user experience.
Patent Information
- Application Number
- CN202511048660.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-29
- Publication Date
- 2025-11-14
AI Technical Summary
Existing question-and-answer processing methods suffer from problems such as large data processing volume, high development costs, and failure to meet user needs in terms of response results.
Build basic data service components for target business scenarios, and generate answer content by matching user questions with components and combining them with a large language model. This includes building a question logic graph and updating logic nodes, and adjusting the material content to meet user needs.
It enables fast and efficient question-and-answer processing, reduces development costs, and improves the accuracy of answers and user experience.
Smart Images
Figure CN120950642A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of data processing technology, and in particular to a question-and-answer processing method, system, electronic device, and storage medium. Background Technology
[0002] In current business scenarios, user questions are generally answered using question-answering models trained with CNN (Convolutional Neural Network) artificial intelligence algorithms, or directly input into existing large language models for processing to output corresponding answers. However, these question-answering methods suffer from problems such as large data processing volume, high development costs, and responses that fail to meet user needs. Summary of the Invention
[0003] The technical problem to be solved by this disclosure is to overcome the above-mentioned defects in the prior art and to provide a question-and-answer processing method, system, electronic device and storage medium.
[0004] This disclosure solves the above-mentioned technical problems through the following technical solution:
[0005] In a first aspect, this disclosure provides a question-and-answer processing method, the question-and-answer processing method comprising:
[0006] Based on several historical business issues in the target business scenario, several categories of basic data service components for answering questions in the target business scenario are constructed.
[0007] Obtain actual user problems in the target business scenario;
[0008] Identify several target basic data service components that match the actual user problem;
[0009] Based on several target basic data service components and a large language model, the target answer content for the actual user question is obtained.
[0010] Optionally, the step of constructing several categories of basic data service components for answering questions in the target business scenario based on several historical business questions in the target business scenario includes:
[0011] Based on several historical business issues in the target business scenario, a problem logic graph is constructed.
[0012] The problem logic graph includes node association information for different logical nodes;
[0013] The node association information includes the node content and the material content of the required answer that matches the node content;
[0014] Based on the node association information of each logical node, corresponding basic data service components are constructed to obtain several categories of basic data service components.
[0015] Optionally, the question-and-answer processing method further includes:
[0016] In response to preset update conditions, the problem logic graph is updated;
[0017] Based on the existence of new logical nodes and corresponding node association information in the problem logic graph, a new basic data service component is generated.
[0018] Optionally, the question-and-answer processing method further includes:
[0019] Based on a preset adjustment scheme, the material content provided by any of the basic data service components is adjusted to form a new basic data service component;
[0020] And / or,
[0021] The complexity of the material content provided by the different categories of basic data service components varies.
[0022] Optionally, before the step of determining a plurality of target basic data service components that match the actual user problem, the method further includes:
[0023] A mapping relationship is pre-built between each of the aforementioned historical business issues and several required basic data service components;
[0024] The step of determining several target basic data service components that match the actual user problem includes:
[0025] The historical business problem with the highest similarity to the actual user problem is calculated, and several basic data service components corresponding to the historical business problem are obtained based on the mapping relationship, so as to serve as several target basic data service components for matching the actual user problem;
[0026] or,
[0027] The step of determining several target basic data service components that match the actual user problem includes:
[0028] Analyze the actual user problem, determine several target logical nodes corresponding to the actual user problem, and obtain several target basic data service components that match the actual user problem.
[0029] Optionally, after the step of determining several target basic data service components that match the actual user problem, the method further includes:
[0030] Obtain the current user's interaction requirements;
[0031] Based on the aforementioned interaction requirements, adjust the currently determined target basic data service components to obtain the adjusted target basic data service components.
[0032] And / or,
[0033] The question-and-answer processing method also includes:
[0034] In response to the existence of new logical nodes in the actual user questions that do not belong to the current question logic graph, a preset construction method is used to construct the node association information corresponding to the new logical nodes to generate new basic data service components, which are then combined with the large language model to obtain the target answer content.
[0035] And / or,
[0036] The step of obtaining the target answer content for the actual user question based on several target basic data service components and the large language model includes:
[0037] Obtain expert experience data that matches the actual user problems;
[0038] The large language model, based on the expert experience data and several target basic data service components, outputs the target answer content for the actual user question.
[0039] A second aspect of this disclosure provides a question-and-answer processing system, the question-and-answer processing system comprising:
[0040] The component building module is used to construct several categories of basic data service components for answering questions in the target business scenario based on several historical business questions in the target business scenario.
[0041] The actual problem acquisition module is used to acquire actual user problems in the target business scenario;
[0042] The target component acquisition module is used to determine several target basic data service components that match the actual user problem;
[0043] The question-and-answer processing module is used to obtain the target answer content corresponding to the actual user question based on several target basic data service components and the large language model.
[0044] Optionally, the target component acquisition module includes:
[0045] The graph construction unit is used to construct a problem logic graph based on several historical business problems in the target business scenario.
[0046] The problem logic graph includes node association information for different logical nodes;
[0047] The node association information includes the node content and the material content of the required answer that matches the node content;
[0048] The target component acquisition unit is used to construct the corresponding basic data service component according to the node association information of each logical node, so as to obtain several categories of basic data service components.
[0049] Optionally, the question-answering processing system further includes:
[0050] The update module is used to update the problem logic graph in response to preset update conditions;
[0051] The component building module is also used to generate a new basic data service component based on the existence of new logical nodes and corresponding node association information in the problem logic graph.
[0052] Optionally, the question-answering processing system further includes:
[0053] The adjustment module is used to adjust the material content provided by any of the basic data service components based on a preset adjustment scheme, so as to form a new basic data service component;
[0054] And / or,
[0055] The complexity of the material content provided by the different categories of basic data service components varies.
[0056] Optionally, the question-answering processing system further includes:
[0057] A relationship building module is used to pre-build a mapping relationship between each of the historical business issues and several required basic data service components;
[0058] The target component acquisition module is further configured to calculate the historical business problem with the highest similarity to the actual user problem, and to acquire several basic data service components corresponding to the historical business problem based on the mapping relationship, so as to serve as several target basic data service components for matching the actual user problem;
[0059] or,
[0060] The target component acquisition module is also used to analyze the actual user problem, determine several target logical nodes corresponding to the actual user problem, and obtain several target basic data service components that match the actual user problem.
[0061] Optionally, the question-answering processing system further includes:
[0062] The requirement elicitation module is used to obtain the current user's interaction requirements;
[0063] The target component acquisition module is further configured to adjust the currently determined target basic data service components based on the interaction requirements, so as to obtain the adjusted target basic data service components.
[0064] And / or,
[0065] The question-and-answer processing system also includes:
[0066] The component construction module is also used to respond to the existence of new logical nodes in the actual user question that do not belong to the current question logic graph, and to construct the node association information corresponding to the new logical node using a preset construction method, so as to generate a new basic data service component, in order to obtain the target answer content in combination with the large language model;
[0067] And / or,
[0068] The question-and-answer processing system also includes:
[0069] The experience data acquisition module is used to acquire expert experience data that matches the actual user problem.
[0070] The question-and-answer processing module is also used to use the large language model based on the expert experience data and several target basic data service components to output the target answer content of the actual user question.
[0071] A third aspect of this disclosure provides an electronic device, including a memory, a processor, and a computer program stored in the memory and for running on the processor, wherein the processor executes the computer program to implement the question-and-answer processing method of the first aspect described above.
[0072] In a fourth aspect, this disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the question-and-answer processing method of the first aspect described above.
[0073] In a fifth aspect, this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the question-and-answer processing method as described in the first aspect above.
[0074] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of this disclosure.
[0075] The positive and progressive effects of this disclosure are as follows:
[0076] In this disclosure, several categories of basic data service components for answering questions in any business scenario are pre-built to enable the rapid, efficient, and automatic matching of one or more corresponding basic data service components for any actual user question. Then, the final answer content is obtained through large language models, etc. This flexibly solves the problems in most business scenarios, saves material production costs, greatly reduces development costs, improves the accuracy of the final question answer, and effectively ensures processing efficiency, thereby enhancing the user experience. Attached Figure Description
[0077] Figure 1 This is a flowchart of the question-and-answer processing method according to Embodiment 1 of this disclosure;
[0078] Figure 2 This is a flowchart of the question-and-answer processing method according to Embodiment 2 of this disclosure;
[0079] Figure 3 This is a schematic diagram of the problem logic diagram of Embodiment 2 of this disclosure;
[0080] Figure 4 This is a schematic diagram of the question-and-answer processing system according to Embodiment 3 of this disclosure;
[0081] Figure 5 This is a schematic diagram of the question-and-answer processing system according to Embodiment 4 of this disclosure;
[0082] Figure 6 This is a schematic diagram of the structure of the electronic device according to Embodiment 5 of this disclosure. Detailed Implementation
[0083] The present disclosure is further illustrated below by way of embodiments, but the present disclosure is not limited to the scope of the embodiments described herein.
[0084] The prefixes such as "first" and "second" used in this disclosure are merely for distinguishing different descriptive objects and do not limit the position, order, priority, quantity, or content of the described objects. The use of ordinal numbers and other prefixes used to distinguish descriptive objects in this disclosure does not constitute a limitation on the described objects. The description of the described objects is given in the claims or the context of the embodiments, and should not be construed as an unnecessary limitation. Furthermore, in the description of this embodiment, unless otherwise stated, "multiple" means two or more.
[0085] In this embodiment of the disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information in specific business scenarios comply with relevant laws and regulations and do not violate public order and good morals.
[0086] Example 1
[0087] like Figure 1 As shown, the question-and-answer processing method in this embodiment includes:
[0088] S101. Based on several historical business questions in the target business scenario, construct several categories of basic data service components for answering questions in the target business scenario;
[0089] Among these, some of the basic data service components fall into several categories and are considered minimum-unit data service components, while others are not. The specific components need to be constructed or adjusted based on the specific business scenario. The target business scenario can be any business scenario, such as service speed detection and analysis for a catering enterprise.
[0090] Historical business issues typically include user questions collected through offline surveys, as well as questions set up based on experience, etc., to form a question set. This question set can be updated regularly or irregularly according to changes over time to ensure the rationality and accuracy of the subsequent determination of basic data service components, thereby ensuring the accuracy and reliability of the final answers to the questions.
[0091] S102. Obtain actual user problems in the target business scenario;
[0092] S103. Identify several target basic data service components that match actual user problems;
[0093] S104. Based on several target basic data service components and a large language model, obtain the target answer content for the actual user's question.
[0094] For example, in step S101, 10 different basic data service components are constructed. For any actual user question A asked by any user in this business scenario, 3 different basic data service components corresponding to the actual user question A are automatically matched and determined. Then, combined with the large language model, the target answer content of the actual user question is obtained quickly and accurately.
[0095] This solution pre-builds several categories of basic data service components for answering questions in any business scenario. This enables the rapid, efficient, and automatic matching of one or more corresponding basic data service components for any actual user question. The final answer is then obtained through large language models, thus flexibly solving most problems in business scenarios. This saves on material production costs, greatly reduces development costs, improves the accuracy of the final question answer, and effectively ensures processing efficiency, thereby enhancing the user experience.
[0096] Example 2
[0097] The question-and-answer processing method in this embodiment is a further improvement on embodiment 1, specifically:
[0098] In a feasible solution, such as Figure 2 As shown, step S101 includes:
[0099] S1011. Based on several historical business issues in the target business scenario, construct a problem logic graph;
[0100] The problem logic graph includes node association information for different logical nodes;
[0101] Node association information includes the node content and the material content of the required answer that matches the node content;
[0102] S1012. Based on the node association information of each logical node, construct the corresponding basic data service components to obtain several categories of basic data service components.
[0103] Specifically, a problem logic graph for the target business scenario can be constructed through pre-defined construction schemes (such as models) and professional personnel analysis. This problem logic graph includes several logical nodes, each logical node corresponding to node content and the material content required for that node content.
[0104] In this solution, a problem logic graph is constructed to determine the logical nodes used to answer questions in the target business scenario and their associated information. This comprehensive and accurate decomposition yields all categories of basic data service components in the target business scenario, ensuring that any question raised by any user in the subsequent actual scenario can be responded to and answered quickly and accurately, thus guaranteeing a good user experience in question-and-answer interaction.
[0105] Taking the target business scenario of monitoring and analyzing service speed in catering enterprises as an example, as shown in Table 1 below, the historical business issues collected for this target business scenario include, but are not limited to:
[0106] Table 1
[0107]
[0108] Based on this, a problem logic graph for the target business scenario is constructed, specifically as follows: Figure 3 As shown;
[0109] Then combine Figure 3 Based on the problem logic graph, the logical nodes, node content, and required material content are determined to construct the basic data service components corresponding to each logical node, as detailed in Table 2 below:
[0110] Table 2
[0111]
[0112] In the table above, parts 1-8 of the basic data service components are the smallest unit data service components, while key components 1-2 are non-smallest unit data service components. Furthermore, the complexity of the material content provided by different categories of basic data service components can be the same or different.
[0113] Specifically, the content provided by the smallest unit of the data service component is less complex, while the content provided by the non-smallest unit of the data service component is more complex.
[0114] The smallest unit of data service components (parts 1-8) provides basic data information and display formats (usually a single piece of information, such as a sales line chart). The non-smallest unit of data service components (key components 1-2) provides the information and display formats required by the main flow and branching points in the problem logic graph (usually the information and formats are more complex, such as an indicator attribution dashboard).
[0115] Referring to Table 3 below, the basic data service components corresponding to each historical business issue in the target business scenario collected in Table 1 are as follows:
[0116] Table 3
[0117]
[0118] For the actual user questions raised by the current user in the target business scenario, one or more corresponding basic data service components are automatically matched, as shown in Table 4 below:
[0119] Table 4
[0120]
[0121] In one feasible solution, the question-and-answer processing method further includes:
[0122] Update the problem logic graph in response to preset update conditions;
[0123] The preset update conditions can be set or adjusted according to actual needs, including but not limited to situations where the market fluctuations in the target business scenario exceed the baseline value.
[0124] Based on the existence of new logical nodes and corresponding node association information in the problem logic graph, new basic data service components are generated.
[0125] In this solution, considering factors such as market fluctuations, the constructed problem logic graph may no longer be accurate and reliable. Therefore, it is necessary to trigger an update of the problem logic graph under certain conditions, and then update it in a timely manner to obtain new basic data service components, so as to further ensure the accuracy and reliability of the answers to actual user questions.
[0126] In one feasible solution, the question-and-answer processing method further includes:
[0127] Based on a preset adjustment scheme, the content of materials provided by any basic data service component is adjusted to form a new basic data service component;
[0128] In this solution, the content provided by each basic data service component can be expanded and adjusted to provide more flexible and reasonable answers to actual user questions, thereby further ensuring that the answers better meet user needs and improving the user's question-and-answer interaction experience.
[0129] Specifically, for example, the material content provided for key components 2, 5, and 6 in Table 2 above can be expanded, that is, newly created and registered. Specifically:
[0130] Key Component 2 - Dashboard - Create new reports, including grid distribution charts showing store service speeds by time period;
[0131] Component 5 - Weather: Create a new intelligent agent that connects to the weather query API (Application Programming Interface) to enable the intelligent agent to input and output the city, date, and weather conditions;
[0132] Part 6 - Establish enterprise knowledge base files, and maintain service speed-related definitions and surrounding promotional information in the knowledge base, etc.
[0133] After adding the necessary materials for the corresponding basic data service components, register all basic data service components (i.e., all parts and key components) on the asset Q&A scheduling platform.
[0134] The specific processing structure for other target business scenarios is similar to the implementation process described above, so it will not be repeated here.
[0135] In a feasible solution, prior to the steps of identifying several target underlying data service components that match the actual user problem, the following are also included:
[0136] Pre-build a mapping relationship between each historical business problem and several required basic data service components;
[0137] Step S103 includes:
[0138] The historical business problem with the highest similarity to the actual user problem is calculated, and several basic data service components corresponding to the historical business problem are obtained based on the mapping relationship, so as to serve as several target basic data service components for matching the actual user problem;
[0139] In this solution, by pre-establishing a mapping relationship between different historical business questions and their corresponding basic data service components, it is possible to quickly determine several target basic data service components for actual user questions, thereby ensuring both the accuracy and efficiency of question-and-answer processing.
[0140] In one feasible embodiment, step S103 includes:
[0141] Analyze actual user problems, identify several target logical nodes corresponding to the actual user problems, and obtain several target basic data service components that match the actual user problems.
[0142] In this solution, by specifically analyzing the actual user questions in the current user query, several target logical nodes involved are automatically identified. In this way, several corresponding target basic data service components are automatically matched, which also effectively ensures the processing efficiency while ensuring the accuracy of question and answer processing.
[0143] In one feasible solution, the steps following step S103 and before step S104 include:
[0144] Obtain the current user's interaction requirements;
[0145] Based on the interaction requirements, adjust the currently determined target basic data service components to obtain the adjusted target basic data service components;
[0146] In this solution, the actual interaction needs of the current user in the target business scenario are considered. For example, if the current user needs a response to the current user's question, the specified basic data service component needs to be considered. If the specified basic data service component is not included in the target basic data service components determined by the above method, then the specified basic data service component needs to be included as one of the target basic data service components. This is to ensure that the subsequent question-and-answer interaction is more in line with the current user's question-and-answer interaction experience, while further enhancing the flexibility and personalization of the question-and-answer interaction scenario.
[0147] In one feasible solution, the question-and-answer processing method further includes:
[0148] In response to the existence of new logical nodes in actual user questions that do not belong to the current question logic graph, a preset construction method is used to construct the node association information corresponding to the new logical nodes, so as to generate new basic data service components, and to obtain the target answer content in combination with the large language model.
[0149] Among them, the preset construction method can be used to automatically generate the corresponding node association information based on model recommendation, etc., which is not limited here, as long as it can be implemented.
[0150] In this solution, under normal circumstances, all logical nodes involved in any user's actual question should be included in the question logic graph, and therefore should be able to match all basic data service components. However, there may be special cases where the logical nodes involved in the user's actual question are partially included in the question logic graph, but one or more new logical nodes are not included in the question logic graph. In this case, it is necessary to promptly construct the node association information corresponding to the new logical nodes to generate new basic data service components, so as to realize timely and effective handling of abnormal situations and further ensure the reliability of the final question answer.
[0151] In one feasible embodiment, step S104 includes:
[0152] S1041. Obtain expert experience data that matches actual user problems;
[0153] S1042. Using a large language model based on expert experience data and several target basic data service components, output the target answer content for actual user questions.
[0154] This solution combines expert experience data to further control the accuracy and reliability of the output content of the large language model, so as to ensure the user's actual question-and-answer interaction needs and user experience.
[0155] Example 3
[0156] like Figure 4 As shown, the question-and-answer processing system in this embodiment includes:
[0157] Component building module 1 is used to build several categories of basic data service components for answering questions in the target business scenario based on several historical business questions in the target business scenario.
[0158] Among these, some of the basic data service components fall into several categories and are considered minimum-unit data service components, while others are not. The specific components need to be constructed or adjusted based on the specific business scenario. The target business scenario can be any business scenario, such as service speed detection and analysis for a catering enterprise.
[0159] Historical business issues typically include user questions collected through offline surveys, as well as questions set up based on experience, etc., to form a question set. This question set can be updated regularly or irregularly according to changes over time to ensure the rationality and accuracy of the subsequent determination of basic data service components, thereby ensuring the accuracy and reliability of the final answers to the questions.
[0160] Module 2 for acquiring actual user problems in the target business scenario;
[0161] Target component acquisition module 3 is used to identify several target basic data service components that match the actual user problem;
[0162] Question and answer processing module 4 is used to obtain the target answer content corresponding to the actual user question based on several target basic data service components and large language model.
[0163] For example, 10 different basic data service components are built through component building module 1. For any actual user question A asked by any user in this business scenario, 3 different basic data service components corresponding to the actual user question A are automatically matched and determined. Then, combined with the large language model, the target answer content of the actual user question is obtained quickly and accurately.
[0164] This solution pre-builds several categories of basic data service components for answering questions in any business scenario. This enables the rapid, efficient, and automatic matching of one or more corresponding basic data service components for any actual user question. The final answer is then obtained through large language models, thus flexibly solving most problems in business scenarios. This saves on material production costs, greatly reduces development costs, improves the accuracy of the final question answer, and effectively ensures processing efficiency, thereby enhancing the user experience.
[0165] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.
[0166] Example 4
[0167] like Figure 5 As shown, the question-and-answer processing system in this embodiment is a further improvement on embodiment 3, specifically:
[0168] In one feasible solution, the target component acquisition module 3 includes:
[0169] Graph construction unit 5 is used to construct a problem logic graph based on several historical business problems in the target business scenario;
[0170] The problem logic graph includes node association information for different logical nodes;
[0171] Node association information includes the node content and the material content of the required answer that matches the node content;
[0172] The target component acquisition unit 6 is used to construct corresponding basic data service components based on the node association information of each logical node, so as to obtain several categories of basic data service components.
[0173] Specifically, a problem logic graph for the target business scenario can be constructed through pre-defined construction schemes (such as models) and professional personnel analysis. This problem logic graph includes several logical nodes, each logical node corresponding to node content and the material content required for that node content.
[0174] In this solution, a problem logic graph is constructed to determine the logical nodes used to answer questions in the target business scenario and their associated information. This comprehensive and accurate decomposition yields all categories of basic data service components in the target business scenario, ensuring that any question raised by any user in the subsequent actual scenario can be responded to and answered quickly and accurately, thus guaranteeing a good user experience in question-and-answer interaction.
[0175] In one feasible solution, the question-answering system also includes:
[0176] Update module 7 is used to update the problem logic graph in response to preset update conditions;
[0177] The preset update conditions can be set or adjusted according to actual needs, including but not limited to situations where the market fluctuations in the target business scenario exceed the baseline value.
[0178] Component building module 1 is also used to generate new basic data service components based on the existence of new logical nodes and corresponding node association information in the problem logic graph.
[0179] In this solution, the content provided by each basic data service component can be expanded and adjusted to provide more flexible and reasonable answers to actual user questions, thereby further ensuring that the answers better meet user needs and improving the user's question-and-answer interaction experience.
[0180] In one feasible solution, the question-answering system also includes:
[0181] Adjustment module 8 is used to adjust the material content provided by any basic data service component based on a preset adjustment scheme to form a new basic data service component;
[0182] And / or,
[0183] The complexity of the material content provided by different categories of basic data service components varies.
[0184] In one feasible solution, the question-answering system also includes:
[0185] Relationship building module 9 is used to pre-build the mapping relationship between each historical business problem and several required basic data service components;
[0186] The target component acquisition module 3 is also used to calculate the historical business problem with the highest similarity to the actual user problem, and to obtain several basic data service components corresponding to the historical business problem based on the mapping relationship, so as to serve as several target basic data service components for matching the actual user problem;
[0187] In this solution, by pre-establishing a mapping relationship between different historical business questions and their corresponding basic data service components, it is possible to quickly determine several target basic data service components for actual user questions, thereby ensuring both the accuracy and efficiency of question-and-answer processing.
[0188] In one feasible solution, the target component acquisition module 3 is also used to analyze actual user problems and determine several target logical nodes corresponding to the actual user problems, so as to obtain several target basic data service components that match the actual user problems.
[0189] In this solution, by specifically analyzing the actual user questions in the current user query, several target logical nodes involved are automatically identified. In this way, several corresponding target basic data service components are automatically matched, which also effectively ensures the processing efficiency while ensuring the accuracy of question and answer processing.
[0190] In one feasible solution, the question-answering system also includes:
[0191] Requirement acquisition module 10 is used to acquire the current user's interaction requirements;
[0192] The target component acquisition module 3 is also used to adjust several currently determined target basic data service components based on interaction requirements, so as to obtain several adjusted target basic data service components;
[0193] In this solution, the actual interaction needs of the current user in the target business scenario are considered. For example, if the current user needs a response to the current user's question, the specified basic data service component needs to be considered. If the specified basic data service component is not included in the target basic data service components determined by the above method, then the specified basic data service component needs to be included as one of the target basic data service components. This is to ensure that the subsequent question-and-answer interaction is more in line with the current user's question-and-answer interaction experience, while further enhancing the flexibility and personalization of the question-and-answer interaction scenario.
[0194] In one feasible solution, the question-answering system also includes:
[0195] The component building module 1 is also used to respond to the existence of new logical nodes in actual user questions that do not belong to the current question logic graph, and to build the node association information corresponding to the new logical nodes using a preset building method, so as to generate new basic data service components, and to obtain the target answer content in combination with the large language model;
[0196] Among them, the preset construction method can be used to automatically generate the corresponding node association information based on model recommendation, etc., which is not limited here, as long as it can be implemented.
[0197] In this solution, under normal circumstances, all logical nodes involved in any user's actual question should be included in the question logic graph, and therefore should be able to match all basic data service components. However, there may be special cases where the logical nodes involved in the user's actual question are partially included in the question logic graph, but one or more new logical nodes are not included in the question logic graph. In this case, it is necessary to promptly construct the node association information corresponding to the new logical nodes to generate new basic data service components, so as to realize timely and effective handling of abnormal situations and further ensure the reliability of the final question answer.
[0198] In one feasible solution, the question-answering system also includes:
[0199] The experience data acquisition module 11 is used to acquire expert experience data that matches actual user problems;
[0200] The question-and-answer processing module 4 is also used to output the target answer content for actual user questions by using a large language model based on expert experience data and several target basic data service components.
[0201] This solution combines expert experience data to further control the accuracy and reliability of the output content of the large language model, so as to ensure the user's actual question-and-answer interaction needs and user experience.
[0202] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.
[0203] Example 5
[0204] Figure 6 This is a schematic diagram of the structure of an electronic device according to an example embodiment of the present disclosure. The electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor. When the processor executes the computer program, it implements the method described in any of the above embodiments. Figure 6 The electronic device 90 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.
[0205] like Figure 6 As shown, the electronic device 90 can be manifested as a general-purpose computing device, such as a server device. The components of the electronic device 90 may include, but are not limited to: at least one processor 91, at least one memory 92, and a bus 93 connecting different system components (including memory 92 and processor 91).
[0206] Bus 93 includes a data bus, an address bus, and a control bus.
[0207] The memory 92 may include volatile memory, such as random access memory (RAM) 921 and / or cache memory 922, and may further include read-only memory (ROM) 923.
[0208] The memory 92 may also include a program tool 925 (or utility) having a set (at least one) program module 924, such program module 924 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.
[0209] The processor 91 executes various functional applications and data processing, such as the methods provided in any of the above embodiments, by running computer programs stored in the memory 92.
[0210] Electronic device 90 can also communicate with one or more external devices 94 (e.g., keyboard, pointing device, etc.). This communication can be performed through input / output (I / O) interface 95. Furthermore, electronic device 90 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public network, such as the Internet) via network adapter 96. As shown, network adapter 96 communicates with other modules of electronic device 90 via bus 93. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 90, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.
[0211] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.
[0212] Example 6
[0213] This disclosure also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in any of the above embodiments.
[0214] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.
[0215] Example 7
[0216] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the above embodiments.
[0217] The program code for executing the computer program product of this disclosure can be written in any combination of one or more programming languages, and the program code can be executed entirely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or entirely on a remote device.
[0218] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.
Claims
1. A question-and-answer processing method, characterized in that, The question-and-answer processing method includes: Based on several historical business issues in the target business scenario, several categories of basic data service components for answering questions in the target business scenario are constructed. Obtain actual user problems in the target business scenario; Identify several target basic data service components that match the actual user problem; Based on several target basic data service components and a large language model, the target answer content for the actual user question is obtained.
2. The question-and-answer processing method as described in claim 1, characterized in that, The step of constructing several categories of basic data service components for answering questions in the target business scenario based on several historical business questions in the target business scenario includes: Based on several historical business issues in the target business scenario, a problem logic graph is constructed. The problem logic graph includes node association information for different logical nodes; The node association information includes the node content and the material content of the required answer that matches the node content; Based on the node association information of each logical node, corresponding basic data service components are constructed to obtain several categories of basic data service components.
3. The question-and-answer processing method as described in claim 2, characterized in that, The question-and-answer processing method also includes: In response to preset update conditions, the problem logic graph is updated; Based on the existence of new logical nodes and corresponding node association information in the problem logic graph, a new basic data service component is generated.
4. The question-and-answer processing method as described in any one of claims 1-3, characterized in that, The question-and-answer processing method also includes: Based on a preset adjustment scheme, the material content provided by any of the basic data service components is adjusted to form a new basic data service component; And / or, The complexity of the material content provided by the different categories of basic data service components varies.
5. The question-and-answer processing method as described in claim 2, characterized in that, Before the step of determining a number of target basic data service components that match the actual user problem, the method further includes: A mapping relationship is pre-built between each of the aforementioned historical business issues and several required basic data service components; The step of determining several target basic data service components that match the actual user problem includes: The historical business problem with the highest similarity to the actual user problem is calculated, and several basic data service components corresponding to the historical business problem are obtained based on the mapping relationship, so as to serve as several target basic data service components for matching the actual user problem; or, The step of determining several target basic data service components that match the actual user problem includes: Analyze the actual user problem, determine several target logical nodes corresponding to the actual user problem, and obtain several target basic data service components that match the actual user problem.
6. The question-and-answer processing method as described in claim 5, characterized in that, Following the step of determining several target basic data service components that match the actual user problem, the method further includes: Obtain the current user's interaction requirements; Based on the aforementioned interaction requirements, adjust the currently determined target basic data service components to obtain the adjusted target basic data service components. And / or, The question-and-answer processing method also includes: In response to the existence of new logical nodes in the actual user questions that do not belong to the current question logic graph, a preset construction method is used to construct the node association information corresponding to the new logical nodes to generate new basic data service components, which are then combined with the large language model to obtain the target answer content. And / or, The step of obtaining the target answer content for the actual user question based on several target basic data service components and the large language model includes: Obtain expert experience data that matches the actual user problems; The large language model, based on the expert experience data and several target basic data service components, outputs the target answer content for the actual user question.
7. A question-and-answer processing system, characterized in that, The question-and-answer processing system includes: The component building module is used to construct several categories of basic data service components for answering questions in the target business scenario based on several historical business questions in the target business scenario. The actual problem acquisition module is used to acquire actual user problems in the target business scenario; The target component acquisition module is used to determine several target basic data service components that match the actual user problem; The question-and-answer processing module is used to obtain the target answer content corresponding to the actual user question based on several target basic data service components and the large language model.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and for running on the processor, characterized in that, When the processor executes the computer program, it implements the question-and-answer processing method according to any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the question-and-answer processing method according to any one of claims 1 to 6.
10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the question-and-answer processing method as described in any one of claims 1 to 6.