Processing method and device, electronic equipment and storage medium

Abstract

The invention discloses a processing method and device, electronic equipment and a storage medium, and relates to the field of computers, in particular to the technical field of artificial intelligence such as large models, deep learning, natural language processing and intelligent search. According to the specific implementation scheme, firstly, a query statement is received, then a reference question and answer pair associated with the query statement is obtained from a reference database, a search result display page is generated according to the reference question and answer pair, and then under the condition that selection operation for any second reply statement is received, an interaction interface associated with any second reply statement is displayed. And finally, updating question and answer pair data associated with any second reply in the reference database based on an interactive operation received in the interactive interface.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, in particular to the field of artificial intelligence technologies such as large models, deep learning, natural language processing, and intelligent search, and specifically to processing methods, devices, electronic devices, and storage media. Background Art

[0002] Currently, search engines typically directly invoke large models for real-time inference and answer generation when users ask questions. This approach relies on the large models' powerful language understanding and generation capabilities, enabling personalized responses to complex and diverse questions. However, due to the computationally intensive and time-consuming inference process of large models, search results are returned slowly, resulting in low processing efficiency. Furthermore, only one large model answer is returned, resulting in insufficiently rich answers and a poor user experience. Summary of the Invention

[0003] The present disclosure provides a processing method, device, electronic device, and storage medium. The specific solution is as follows:

[0004] According to one aspect of the present disclosure, a processing method is provided, comprising:

[0005] Receive query statements;

[0006] Acquire a reference question-answer pair associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first answer statement;

[0007] generating a search result display page based on the reference question-answer pair, wherein the length of the second answer sentence in the target question-answer pair on the search result display page is less than the length of the corresponding first answer sentence;

[0008] When a selection operation is received for any second reply statement, an interactive interface associated with the second reply statement is displayed, wherein the interactive interface is an interface for generating the corresponding first reply statement;

[0009] Based on the interactive operation received in the interactive interface, the question-answer pair data associated with any second answer in the reference database is updated.

[0010] According to another aspect of the present disclosure, there is provided a processing device, comprising:

[0011] A receiving module, used for receiving a query statement;

[0012] an acquisition module, configured to acquire a reference question-answer pair associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first answer statement;

[0013] a generating module for generating a search result display page based on the reference question-answer pair, wherein the length of the second answer statement in the target question-answer pair on the search result display page is less than the length of the corresponding first answer statement;

[0014] A display module, configured to, upon receiving a selection operation for any second reply statement, display an interactive interface associated with the second reply statement, wherein the interactive interface is an interface for generating the corresponding first reply statement;

[0015] An updating module is used to update the question-answer pair data associated with any second reply in the reference database based on the interactive operation received in the interactive interface.

[0016] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0017] at least one processor; and

[0018] a memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the above embodiment.

[0020] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method according to the above embodiment.

[0021] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which implements the steps of the method described in the above embodiment when executed by a processor.

[0022] The processing method, device, electronic device and storage medium provided by the present disclosure have the following beneficial effects: first, a query statement is received, and then a reference question-answer pair associated with the query statement is obtained from a reference database, and a search result display page is generated based on the reference question-answer pair. After that, when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Finally, based on the interactive operation received in the interactive interface, the question-answer pair data associated with any second reply in the reference database is updated. Thus, by obtaining the associated question-answer pair data from the reference database based on the query statement, and generating a search result display page based on the question-answer pair data, the search engine can be effectively prevented from repeatedly calling a large model to generate answers for the same or similar questions, thereby shortening the processing time, improving the utilization rate of computing resources and processing efficiency, and improving the richness of search results. When a selection operation is received for any question-answer pair, an interactive interface associated with the corresponding question-answer pair is displayed, and based on the interactive operation of the user in the interactive interface, the corresponding question-answer pair data in the reference database is updated, so that the data in the database can be updated in a timely manner, the reliability and accuracy of the search results can be improved, and the user experience can be enhanced.

[0023] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0025] Figure 1 A flowchart of a processing method provided in one embodiment of the present disclosure;

[0026] Figure 2 A schematic diagram of a search result display page and an interactive interface generated by the processing method proposed in an embodiment of the present disclosure;

[0027] Figure 3 A flowchart of a processing method provided in another embodiment of the present disclosure;

[0028] Figure 4 A flowchart of a processing method provided in another embodiment of the present disclosure;

[0029] Figure 5 A flowchart of a processing method provided in another embodiment of the present disclosure;

[0030] Figure 6 A flowchart of a processing method provided in another embodiment of the present disclosure;

[0031] Figure 7 A schematic diagram of the overall implementation logic of the processing method proposed in the embodiment of the present disclosure;

[0032] Figure 8 A schematic diagram of the structure of a processing device provided in one embodiment of the present disclosure;

[0033] Figure 9 It is a block diagram of an electronic device used to implement the processing method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0034] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0035] The embodiments of the present disclosure relate to artificial intelligence technology fields such as large models, deep learning, natural language processing, and intelligent search.

[0036] Artificial Intelligence (AI) is a new technical science that studies and develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence.

[0037] The large model can also be called the Foundation Model. The model extracts knowledge from billions of corpora or images, learns, and then produces a large model with billions of parameters.

[0038] Deep learning (DL) involves learning the inherent patterns and representational hierarchies of sample data. The information gained from this learning process is highly helpful in interpreting data such as text, images, and sounds. The ultimate goal of deep learning is to enable machines to have the same analytical and learning capabilities as humans, enabling them to recognize data such as text, images, and sounds.

[0039] Natural Language Processing (NLP) is an interdisciplinary subject in the fields of computer science, artificial intelligence, and linguistics. It mainly studies how to enable computers to understand, process, generate, and simulate human language, so as to achieve the ability to have natural conversations with humans.

[0040] Intelligent Search is an advanced search service based on artificial intelligence and big data technologies, designed to improve user efficiency and experience in information retrieval through intelligent means. It not only supports traditional keyword matching but also incorporates advanced technologies such as natural language processing, machine learning, and vector search to understand user query intent and provide more accurate and diverse search results. Intelligent Search analyzes user query intent and supports features such as synonym recognition, spelling correction, and word segmentation analysis.

[0041] It should be noted that the acquisition, storage, use, and processing of data in the disclosed technical solution comply with the relevant provisions of national laws and regulations and do not violate public order and good morals.

[0042] The processing method, apparatus, electronic device, and storage medium according to the embodiments of the present disclosure are described below with reference to the accompanying drawings.

[0043] Figure 1 A flowchart of a processing method provided in one embodiment of the present disclosure.

[0044] like Figure 1 As shown, the processing method includes:

[0045] Step 101: Receive a query statement.

[0046] It should be noted that the query statement can be input by the user in the search engine and can be determined according to the user's needs. For example, the query statement can be "What is the principle of large model A?", etc., which is not limited in this disclosure.

[0047] In the present disclosure, a query statement may be received through a search engine.

[0048] Step 102: Obtain reference question-answer pairs associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first reply statement.

[0049] It should be noted that the reference database may be a question-answer pair database for storing historical question-answer pair data, which may be pre-built. In other words, each of the multiple question-answer pair data included in the reference database is a question-answer pair data that has been generated by the search engine.

[0050] In the present disclosure, the question-answer pair data in the reference database includes a question statement and a generated first answer statement. In other words, the reference database includes multiple question-answer pair data, each of which includes a question statement and a generated first answer statement.

[0051] It should be noted that the first answer statement of the question-answer pair data in the reference database can be generated based on the question statement using a large model. In other words, the reference database includes multiple question-answer pairs, and the first answer statement included in each question-answer pair can be generated based on the question statement in the corresponding question-answer pair data using a large model.

[0052] The specific type and structure of the large model can be pre-set as needed. For example, the large model can be a Transformer-based generative model, a multimodal fusion model, or a generative question-answering system, or a retrieval-enhanced generative architecture, etc., and this disclosure does not limit this.

[0053] In some possible implementations, when constructing a reference database, the question statement included in each question-answer pair in the reference database can be used as the index corresponding to the question-answer pair, thereby improving the efficiency of retrieval and matching in the reference database.

[0054] In some possible implementations, each question-and-answer pair may also include user feedback data, which can reflect the user's approval of the question-and-answer pair. For example, the feedback data may include user ratings and comments on the question-and-answer pair, as well as user actions (such as clicks) on the question-and-answer pair. This can assist search engines in sorting multiple reference question-and-answer pairs, enabling users to quickly locate high-quality answers.

[0055] Among them, the reference question-answer pair can be composed of a question statement with a high similarity to the query statement and its corresponding first answer statement. The number of the question-answer pair can be determined according to the actual situation. It can be one or more, and this disclosure does not limit this.

[0056] In some possible implementations, when obtaining reference question-answer pairs associated with a query statement from a reference database, the reference question-answer pairs associated with the query statement can be determined by determining the similarity between the question statement of each question-answer pair in the reference database and the query statement. In other words, a question statement and its corresponding first answer statement with a high similarity can be determined as a reference question-answer pair, although this disclosure is not limited to this.

[0057] It should be noted that the specific algorithm for determining the similarity between the question statement and the query statement in each question-answer pair in the reference database can be pre-set or determined as needed. For example, a similarity algorithm such as cosine similarity or Jaccard similarity can be used to determine the similarity between the question statement and the query statement in each question-answer pair, but this disclosure does not limit this.

[0058] Step 103: Generate a search result display page based on the reference question and answer pair, wherein the length of the second answer statement in the target question and answer pair on the search result display page is less than the length of the corresponding first answer statement.

[0059] It should be noted that the number of search result display pages can be determined based on the number of reference question-answer pairs and the number of question-answer pairs displayed on each page. In other words, when generating a search result display page based on the reference question-answer pairs, the number of search result display pages can be one page, or multiple pages, each of which can include multiple question-answer pairs. Since the length of the second answer statement in the target question-answer pair on the search result display page is shorter than the length of the corresponding first answer statement, the efficiency of users in obtaining information can be improved, thereby enhancing the user experience.

[0060] In some possible implementations, when generating a search result display page based on a reference question-and-answer pair, any of the following may be included: generating a second answer statement on the search result display page based on the summary information corresponding to the first answer statement in the reference question-and-answer pair; or generating the second answer statement on the search result display page based on a portion of the first answer statement in the reference question-and-answer pair. Generating the second answer statement on the search result display page based on the summary information or a portion of the first answer statement effectively ensures the conciseness of the information on the search result display page, helps users quickly determine the relevance of the answer statements, significantly saves time, and thus improves the user experience.

[0061] It should be noted that the present disclosure can determine the summary information corresponding to the first reply statement through any preset algorithm. For example, the key content in the first reply statement can be directly extracted through an extractive method, such as using the TextRank algorithm, the Term Frequency-Inverse Document Frequency (TF-IDF) algorithm, etc. Alternatively, summary information can be generated based on the first reply statement through an abstract method, such as using a Sequence to Sequence (Seq2Seq) model, etc., which is not limited by the present disclosure.

[0062] Among them, some sentences in the first reply sentence can be the first few sentences in the first reply sentence (such as the first two sentences, which can be set according to needs), or can also be the front and back sentences in the first reply sentence (such as the first two sentences and the last two sentences, which can be set according to needs), or can also be the key sentences obtained after identifying the first reply sentence, etc., and the present disclosure does not limit this.

[0063] In some possible implementations, the search result display page may also include at least one of the following: the matching degree between the question statement and the query statement in the target question and answer pair; the confidence of the first reply statement corresponding to the target question and answer pair; the generation time information of the target question and answer pair; and the model information for generating the corresponding first reply statement. This at least one item of information can help users quickly judge the reliability and relevance of the target question and answer pair.

[0064] It should be noted that the degree of match between the question statement and the query statement in the target question-answer pair can be determined by a preset matching degree calculation method. For example, it can be determined by a text similarity calculation method such as cosine similarity, Jaccard similarity, etc., which is not limited in this disclosure.

[0065] Among them, the confidence of the first reply statement can be used to represent the credibility of the first reply statement, and can also be considered as user recognition, which is not limited in this disclosure.

[0066] It should be noted that the confidence of the first reply statement in this disclosure can be obtained by weighted calculation based on multi-dimensional feedback data such as users' comments, ratings, and selected operations (such as clicks) on the first reply statement, and this disclosure does not limit this.

[0067] Among them, the model information for generating the corresponding first reply statement may include the model version for generating the first reply statement, the algorithm associated with the model, etc., which is not limited in this disclosure.

[0068] Step 104: upon receiving a selection operation for any second reply statement, displaying an interactive interface associated with any second reply statement.

[0069] Among them, the interactive interface is an interface for generating the corresponding first reply statement.

[0070] The selection operation may be any preset operation, for example, a click operation, etc., which is not limited in the present disclosure.

[0071] In the present disclosure, the first reply statement is generated by the macro model based on the question statement. Therefore, the interactive interface can be a dialog interface for the macro model to generate the first reply statement. This interface can include the question statement, the first reply statement, the references used by the macro model to generate the first reply statement, and interactive components such as a dialog box for user input, buttons for users to like or dislike, controls for users to rate, or comment boxes, etc., which are not limited in the present disclosure.

[0072] That is to say, after displaying the interactive interface associated with any second reply statement, the user can conduct further inquiries in the interactive interface, or can also rate or comment on the first reply statement, etc., and this disclosure does not limit this.

[0073] Below Figure 2 Taking as an example, the search result display page and interactive interface generated based on the processing method disclosed in the present invention are illustrated. Figure 2 This is a schematic diagram of a search result display page and an interactive interface generated in the processing method proposed in an embodiment of the present disclosure.

[0074] like Figure 2 As shown, the query statement is "What is the principle of large model A?", and when searching in the reference database based on the query statement, four reference question-answer pairs are obtained, namely, "What is the principle of large model A?", "What is the working principle of large model A?", and "What is the working principle of large model A?", wherein the number of reference question-answer pairs is only an example here and is not limited. Then, based on these four reference question-answer pairs, a search result display page can be generated, and the question statement of each question-answer pair, the corresponding second answer statement, the matching degree between the question statement and the query statement, the confidence of the first answer statement in each question-answer pair, and the question-answer pair generation time information are displayed on the search result display page. Among them, the display content and layout in the operation result display page can be set as needed. This is only an example here and is not limited.

[0075] When a user selects the search result for the first question-answer pair, "What is the principle behind large model A?", an interactive interface for generating the first answer statement for the question-answer pair can be opened, and the corresponding first answer statement and the references used when generating the first answer statement can be displayed to the user through the interactive interface. The user can also conduct further inquiries by entering a question in the interactive interface, or can also rate the current first answer statement. The display content, interactive components, and layout in the interactive interface can be configured as needed. This is only an example and is not intended to be limiting.

[0076] Step 105: Based on the interactive operation received in the interactive interface, update the question-answer pair data associated with any second answer in the reference database.

[0077] Among them, interactive operations may include any operations such as liking, scoring, and evaluating the first reply statement in the interactive interface, and this disclosure does not limit this.

[0078] In the present disclosure, after displaying an interactive interface associated with any second reply statement, upon receiving an interactive operation from a user in the interactive interface, the question-answer pair data associated with the second reply statement in the reference database may be updated based on the interactive operation. For example, if the interactive operation is a scoring operation, the confidence level of the question-answer pair data associated with the second reply statement may be updated based on the user's score, and the like, but the present disclosure does not impose any limitation on this.

[0079] In the disclosed embodiment, a query statement is first received, and then a reference question-answer pair associated with the query statement is obtained from a reference database, and a search result display page is generated based on the reference question-answer pair. After that, when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Finally, based on the interactive operation received in the interactive interface, the question-answer pair data associated with any second reply in the reference database is updated. Thus, by obtaining the associated question-answer pair data from the reference database based on the query statement, and generating a search result display page based on the question-answer pair data, the search engine can be effectively prevented from repeatedly calling a large model to generate answers for the same or similar questions, thereby shortening the processing time, improving the utilization rate of computing resources and processing efficiency, and improving the richness of search results. When a selection operation is received for any question-answer pair, an interactive interface associated with the corresponding question-answer pair is displayed, and based on the interactive operation of the user in the interactive interface, the corresponding question-answer pair data in the reference database is updated, so that the data in the database can be updated in a timely manner, the reliability and accuracy of the search results can be improved, and the user experience can be enhanced.

[0080] Figure 3 A flowchart of a processing method provided in another embodiment of the present disclosure.

[0081] like Figure 3 As shown, the processing method includes:

[0082] Step 301: Receive a query statement.

[0083] Step 302: Obtain reference question-answer pairs associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first reply statement.

[0084] The specific implementation of steps 301 to 302 can be referred to the detailed descriptions in other embodiments of the present disclosure, and will not be described in detail here.

[0085] Step 303: Determine the display order of the target question-answer pairs in the search result display page based on the matching degree between the question statement and the query statement in the reference question-answer pair and the confidence level of the corresponding first reply statement.

[0086] In the present disclosure, after obtaining the reference question-answer pairs associated with the query statement, before generating the search result display page, the display order of the target question-answer pairs in the search result display page can be determined first based on the matching degree between the question statement and the query statement in the reference question-answer pair, and the confidence of the corresponding first reply statement. In other words, for each target question-answer pair in the search result display page, its display order in the search result display page can be determined based on the matching degree between the question statement and the query statement in the corresponding reference question-answer pair, and the confidence of the corresponding first reply statement. In this way, the target question-answer pairs with higher matching degree and higher confidence level can be displayed in an earlier order, so that users can quickly obtain high-quality answers and improve user experience. The present disclosure does not limit this.

[0087] Step 304: Generate a search result display page based on the reference question-answer pair and the display order of the target question-answer pair in the search result display page, wherein the length of the second answer statement in the target question-answer pair in the search result display page is less than the length of the corresponding first answer statement.

[0088] In the present disclosure, after determining the display order of the target question-answer pairs in the search result display page, a search result display page can be generated based on the reference question-answer pairs and the display order, thereby improving the quality and reliability of the search results and providing conditions for improving the user experience.

[0089] In some possible implementations, when a search results display page is generated based on reference question-answer pairs and the display order of target question-answer pairs in the search results display page, the target question-answer pairs can be identified in the search results page. For example, target question-answer pairs with high matching or confidence levels can be highlighted to help users quickly locate high-quality answers. This disclosure does not limit this.

[0090] Step 305: upon receiving a selection operation for any second reply statement, displaying an interactive interface associated with any second reply statement, wherein the interactive interface is an interface for generating a corresponding first reply statement.

[0091] Step 306: Based on the interactive operation received in the interactive interface, update the question-answer pair data associated with any second answer in the reference database.

[0092] The specific implementation of steps 304 to 306 can refer to the detailed descriptions in other embodiments of the present disclosure and will not be described in detail here.

[0093] In an embodiment of the present disclosure, a query statement is first received, and then a reference question-answer pair associated with the query statement is obtained from a reference database. The display order of the target question-answer pairs in the search result display page is determined based on the matching degree between the question statement and the query statement in the reference question-answer pair and the confidence of the corresponding first reply statement. Thereafter, a search result display page is generated based on the reference question-answer pairs and the display order of the target question-answer pairs in the search result display page. When a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Finally, based on the interactive operation received on the interactive interface, the question-answer pair data associated with any second reply in the reference database is updated. Therefore, after obtaining the question-answer pairs associated with the query statement from the reference database, before generating the search result display page, the display order of the question-answer pairs in the search result display page can be determined based on the matching degree between the question in the question-answer pair and the query statement, and the confidence of the corresponding reply statement, and the search result display page can be generated based on the question-answer pairs and the display order, so that users can quickly locate high-quality answers, improve the effect and quality of search results, and enhance user experience. When a selection operation is received for any question-answer pair, its associated interactive interface is displayed, and based on the user's interactive operation in the interactive interface, the corresponding question-answer pair data in the database is updated, thereby improving the quality and accuracy of the information in the database and improving the reliability and effect of the processing method.

[0094] Figure 4 A flowchart of a processing method provided in another embodiment of the present disclosure.

[0095] like Figure 4 As shown, the processing method includes:

[0096] Step 401: Receive a query statement.

[0097] Step 402: Obtain reference question-answer pairs associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first reply statement.

[0098] Step 403: Generate a search result display page based on the reference question and answer pair, wherein the length of the second answer statement in the target question and answer pair on the search result display page is less than the length of the corresponding first answer statement.

[0099] Step 404, when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed, wherein the interactive interface is an interface for generating the corresponding first reply statement.

[0100] The specific implementation of steps 401 to 404 can refer to the detailed descriptions in other embodiments of the present disclosure and will not be described in detail here.

[0101] Step 405: When the interactive operation received by the interactive interface is a feedback operation for the corresponding first reply statement, the confidence of the corresponding first reply statement is updated based on the type and / or emotional direction of the feedback operation.

[0102] Among them, the types of feedback operations may include likes, comments, and ratings, etc., which are not limited in this disclosure.

[0103] The emotional direction of the feedback operation can be determined based on the specific feedback operation. For example, when the feedback operation is a like, the emotional direction is positive, and when the feedback operation is a dislike, the emotional direction is negative. When the feedback operation is a rating, the emotional direction can be determined based on a quantitative form such as a score, etc. This disclosure does not limit this.

[0104] It should be noted that the specific update of the confidence of the corresponding first reply statement can be determined according to the type of feedback operation and / or the emotional direction. For example, when the type of feedback operation is a like, it can be determined that the user approves the corresponding first reply statement, and the confidence of the first reply statement can be improved. Alternatively, when the type of feedback operation is an evaluation and its emotional direction is negative, it can be determined that the quality of the corresponding first reply statement is poor and has not been approved by the user. In this case, the confidence of the first reply statement can be reduced, and so on. This disclosure does not limit this.

[0105] Step 406: When the interactive operation received by the interactive interface is an operation on an associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation.

[0106] The reference information source may be any type of information source used to generate the first reply statement. For example, the reference information source may include documents, web pages, multimedia information sources such as images or videos, knowledge graphs, and expert knowledge, etc., which are not limited in this disclosure.

[0107] It should be noted that the operation on the associated reference information source may be clicking and reading or viewing the reference information source, or evaluating the reference information source, etc., and this disclosure does not limit this.

[0108] The operation duration may be the duration that the user stays on the reference information source, such as the duration that the user reads or views a reference information source, etc., which is not limited in the present disclosure.

[0109] It should be noted that the specific update operation of the weight of the information source associated with the corresponding first reply statement can be determined according to the type, duration and / or emotional direction of the operation on the associated reference information source. For example, when the user clicks and reads multiple information sources associated with the first reply statement, the weights of these multiple information sources can be updated according to the reading time of each of these multiple information sources. When the reading time of the information source is longer, it can be determined that the reliability of the corresponding information source is higher, and the user's recognition of the corresponding information source is higher. At this time, the weight of the corresponding information source can be increased. Conversely, when the reading time of the information source is shorter, it can be determined that the user's recognition of the corresponding information source is lower. At this time, the weight of the corresponding information source can be reduced, etc. Or when the user makes a positive evaluation of any information source, it can be determined that the user has recognized any information source. At this time, the weight of the corresponding information source can be increased, etc. This disclosure does not limit this.

[0110] In some possible implementation forms, after the weight of the information source associated with the corresponding first reply statement is updated, the first reply statement can be regenerated based on the updated weight, thereby improving the quality and accuracy of the first reply statement. This disclosure does not limit this.

[0111] In some possible implementation forms, after the corresponding question-answer pair data is updated based on the interactive operation, the relevant data in the reference database can also be updated based on the updated question-answer pair data, thereby improving the data quality in the reference database.

[0112] Among them, when the question and answer pair data associated with any second reply in the reference database is updated based on the interactive operation received in the interactive interface, when the interactive operation includes a feedback operation on the corresponding first reply statement and an operation on the associated reference information source, the specific updating method of the corresponding question and answer pair data may include the specific implementation form of steps 405 and 406.

[0113] Alternatively, when the interactive operation is a feedback operation for the corresponding first reply statement, the specific updating method may only include the specific implementation form of step 405.

[0114] Alternatively, when the interactive operation is an operation on an associated reference information source, the specific updating method may only include the specific implementation form of step 406, which is not limited in this disclosure.

[0115] That is to say, in the case where the interactive operation includes a feedback operation on the corresponding first reply statement and an operation on the associated reference information source, the processing method proposed in the embodiment of the present disclosure may include specific implementation forms of steps 401 to 406. At this time, according to the specific implementation forms of steps 405 and 406, when updating the question-answer pair data in the reference database, not only can the confidence of the corresponding first reply statement be updated based on the type and / or emotional direction of the feedback operation, but the weight of the information source associated with the first reply statement can also be updated based on the type, duration and / or emotional direction of the operation on the associated reference information source.

[0116] Alternatively, when the interactive operation is a feedback operation for the corresponding first reply statement, the processing method proposed in the embodiment of the present disclosure may include specific implementations of steps 401 to 405. In this case, according to the specific implementation of step 405, when updating the question-answer pair data in the reference database, the confidence level of the corresponding first reply statement may be updated based on the type and / or sentiment direction of the feedback operation.

[0117] Alternatively, when the interactive operation is directed to an associated reference information source, the processing method proposed in the embodiment of the present disclosure may include specific implementations of steps 401 to 404 and step 406. In this case, depending on the specific implementation of step 406, when updating the question-answer pair data in the reference database, the weight of the information source associated with the first reply statement may be updated based on the type, duration, and / or sentiment direction of the operation directed to the associated reference information source, although this disclosure does not limit this.

[0118] It should be noted that, when the processing method proposed in the embodiments of the present disclosure includes specific implementations of steps 401 to 406, when updating the question-answer pair data in the reference database according to the specific implementations of steps 405 and 406, the execution order of steps 405 and 406 can be determined as needed. For example, steps 405 and 406 can be executed in parallel, or step 405 can be executed first and then step 406 (or step 405 first and then step 406), and this disclosure does not limit this.

[0119] In the embodiment of the present disclosure, a query statement is first received, and then a reference question-answer pair associated with the query statement is obtained from a reference database, and a search result display page is generated based on the reference question-answer pair. After that, when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Finally, when the interactive operation received on the interactive interface is a feedback operation for the corresponding first reply statement, the confidence of the corresponding first reply statement is updated based on the type and / or emotional direction of the feedback operation. When the interactive operation received on the interactive interface is an operation for an associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation. Thus, after displaying the interactive interface associated with the question-answer pair selected by the user, the confidence of the corresponding reply statement and / or the weight of the associated information source are updated based on the interactive operation, providing conditions for improving the processing effect and the data quality in the database.

[0120] Figure 5 A flowchart of a processing method provided in another embodiment of the present disclosure.

[0121] like Figure 5 As shown, the processing method includes:

[0122] Step 501: Receive a query statement.

[0123] Step 502: Obtain reference question-answer pairs associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first reply statement.

[0124] Step 503: Generate a search result display page based on the reference question and answer pair, wherein the length of the second answer statement in the target question and answer pair on the search result display page is less than the length of the corresponding first answer statement.

[0125] Step 504: upon receiving a selection operation for any second reply statement, displaying an interactive interface associated with any second reply statement, wherein the interactive interface is an interface for generating a corresponding first reply statement.

[0126] Step 505: When the interactive operation received by the interactive interface is a feedback operation for the corresponding first reply statement, the confidence of the corresponding first reply statement is updated based on the type and / or emotional direction of the feedback operation.

[0127] Step 506: When the interactive operation received by the interactive interface is an operation on an associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation.

[0128] The specific implementation of steps 501 to 506 can refer to the detailed descriptions in other embodiments of the present disclosure and will not be described in detail here.

[0129] Step 507: When the confidence corresponding to any first reply statement is less than a threshold, or when the weight of the reference information source associated with any first reply statement is updated, the first question statement corresponding to any first reply statement is marked as a task to be run.

[0130] The threshold value may be a confidence critical value used to judge the credibility of any first reply statement, which may be pre-set or determined as needed, and the present disclosure does not limit this.

[0131] Among them, the tasks to be run in batches can be automated tasks for regenerating reply statements for large models.

[0132] In the present disclosure, after the confidence of any first reply statement is updated, and / or the weight of the information source associated with it is updated, if the confidence corresponding to the any first reply statement is less than a threshold value, it can be determined that the credibility of the any first reply statement is low and the quality of the any first reply statement is poor. At this time, the first question statement corresponding to the any first reply statement can be marked as a task to be run in batches, waiting for the large model to regenerate a new reply statement. Alternatively, in the case where the weight of the reference information source associated with the any reply statement is updated, it can be determined that the user's recognition of the associated reference information source has changed. At this time, in order to improve the reliability of the any first reply statement, the first question statement corresponding to the any first reply statement can be marked as a task to be run in batches.

[0133] In some possible implementation forms, when marking the first question statement corresponding to any first reply statement as a task to be run in batches, the reference database can also be screened, and the question statements corresponding to other first reply statements in the reference database whose confidence level is less than a threshold are marked as tasks to be run in batches. This disclosure does not limit this.

[0134] Step 508: When the batch task execution conditions are met, execute the task to be batched to generate a new answer statement corresponding to the first question statement.

[0135] In some possible implementations, the execution condition for a batch task may include at least one of the following: the amount of remaining computing resources in the system is greater than a resource threshold; a preset time period; or the number of pending batch tasks is greater than a threshold. This at least one condition ensures the quality and effectiveness of the system's execution of pending batch tasks.

[0136] Among them, the resource threshold can be used to determine whether there are sufficient computing resources in the system to execute the critical value of the remaining computing resources when the batch of tasks to be run is to be executed. It can be set as needed, and this disclosure does not limit this.

[0137] The preset time period may be a time period during which the system executes the pending batch of tasks. To ensure that the system has sufficient computing resources to execute the pending batch of tasks, the preset time period may be set to a time period during which the system's computing resources are relatively idle. For example, the preset time period may be evening, etc., and this disclosure does not limit this.

[0138] Among them, the quantity threshold can be a critical value of the number of tasks to be run for judging whether the system can start executing the tasks to be run. It can be pre-set or determined according to actual needs. This disclosure does not limit this.

[0139] In the present disclosure, in order to improve resource utilization and avoid the system executing tasks to be run when the number of tasks to be run is too small, resulting in resource waste, tasks to be run can be executed only when the number of tasks to be run is greater than a threshold.

[0140] In the present disclosure, after determining the tasks to be run in batches, the tasks to be run in batches are executed when the execution conditions of the tasks are met, thereby generating new reply statements corresponding to the question statements, improving the quality of the reply statements, and improving the reliability and effect of the processing method.

[0141] In some possible implementations, after generating a new response statement corresponding to the question statement, the new response statement can be evaluated for quality, for example, by using indicators such as language fluency and information accuracy to ensure that the quality of the new response statement is improved.

[0142] In some possible implementations, after generating a new answer statement corresponding to a question statement, the question statement and its corresponding new answer statement may be associated and stored in a reference database, thereby improving the data quality and reliability in the reference database.

[0143] In an embodiment of the present disclosure, a query statement is first received, and a reference question-answer pair associated with the query statement is obtained from a reference database. Then, based on the reference question-answer pair, a search result display page is generated, and when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Then, when the interactive operation received on the interactive interface is a feedback operation for the corresponding first reply statement, the confidence of the corresponding first reply statement is updated based on the type and / or emotional direction of the feedback operation. When the interactive operation received on the interactive interface is an operation for the associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation. Finally, when the confidence corresponding to any first reply statement is less than a threshold, or when the weight of the reference information source associated with any first reply statement is updated, the first question statement corresponding to any first reply statement is marked as a task to be batch run, and when the batch task execution conditions are met, the task to be batch run is executed to generate a new reply statement corresponding to the first question statement. Therefore, after the confidence of the corresponding reply statement and / or the weight of the associated information source is updated based on the user's interactive operation in the interactive interface, the question statement corresponding to the reply statement with lower confidence, or the question statement corresponding to the reply statement with updated weight of the associated information source, is determined as a task to be batched, and when the execution conditions of the batch task are met, the task to be batched is executed to generate a corresponding new reply statement, thereby improving the quality of the question and answer pairs and improving the efficiency and reliability of the processing method.

[0144] Figure 6 A flowchart of a processing method provided in another embodiment of the present disclosure.

[0145] like Figure 6 As shown, the processing method includes:

[0146] Step 601: Receive a query statement.

[0147] The specific implementation of step 601 can refer to the detailed descriptions in other embodiments of the present disclosure and will not be described in detail here.

[0148] Step 602: If the reference database does not contain a reference question-answer pair associated with the query statement, the query statement is marked as a task to be run.

[0149] In the present disclosure, after receiving a query statement, the reference database can be first searched based on the query statement. If the reference database does not contain a reference question-answer pair associated with the query statement, the query statement can be marked as a task to be run in batches. Then, the large model can be used to generate a reply statement corresponding to the query statement in real time, and the reply statement can be returned to the user. The present disclosure does not limit this.

[0150] In some possible implementation forms, after using a large model to generate a reply statement corresponding to a query statement in real time, the query statement and the reply statement can also be associated and stored in a reference database to enrich the data in the reference database. This disclosure does not limit this.

[0151] Step 603: When the batch task execution conditions are met, the batch task to be run is executed to generate a reply statement corresponding to the query statement.

[0152] In some possible implementation forms, after executing the batch task to be run and generating the response statement corresponding to the query statement, the query statement and the generated response statement can be associated and stored in the reference database, thereby improving the richness of the question-answer pairs in the reference database.

[0153] The specific implementation of step 603 can be referred to the detailed description in other embodiments of the present disclosure, and will not be described in detail here.

[0154] In the disclosed embodiment, a query statement is first received. If a reference database does not contain a reference question-answer pair associated with the query statement, the query statement is marked as a pending task. Finally, if the execution conditions for the batch task are met, the pending task is executed to generate a corresponding reply statement for the query statement. Thus, if a reference database does not contain a reference question-answer pair associated with the query statement, the query statement can be marked as a pending task. If the execution conditions for the batch task are met, the pending task is executed to generate a corresponding reply statement for the query statement, thereby improving the reliability of the processing method.

[0155] The following combination Figure 7 , the implementation logic of the processing method proposed in the embodiment of the present disclosure is explained. Figure 7 This is a schematic diagram of the overall implementation logic of the processing method proposed in the embodiment of the present disclosure.

[0156] Figure 7 In the embodiment of the present disclosure, the overall implementation logic architecture of the processing method proposed includes a question and answer storage and retrieval module, a reference database, a large language model, a user feedback optimization module, and a batch running module.

[0157] The question-answer storage and retrieval module is used to establish a reference database to store historical question-answer pair data, build an index for the question-answer pair data in the reference database, achieve fast retrieval and matching, and retrieve the reference database based on the received query statement;

[0158] A reference database, used to store historical question-answer pairs generated by the large model, which may include question statements, answer statements, generation time, confidence level, and other data;

[0159] A large language model is used to generate the corresponding answer sentence for the question sentence;

[0160] The user feedback optimization module is used to update the corresponding question-answer data according to the user's interactive operations, and update the relevant data in the reference database;

[0161] The batch running module is used to execute the batch running task and generate new response statements corresponding to the question statements and / or query statements based on the question statements corresponding to the low-confidence response statements in the reference database and / or the query statements for which similar questions have not been retrieved, when the conditions for executing the batch running task are met.

[0162] like Figure 7 As shown, after receiving the query statement input by the user, the reference database is retrieved through the question and answer storage and retrieval module based on the query statement. When a question similar to the query statement is retrieved in the reference database, the reference question and answer pair associated with the query statement is returned, and based on the reference question and answer pair, a search result display page is generated to display the question and answer pair results to the user.

[0163] When no questions similar to the query statement are retrieved in the reference database, a large language model is used to generate a reply statement corresponding to the query statement in real time, and the generated reply statement is returned to the user for display, and the generated reply statement is associated with the query statement and stored in the database.

[0164] Furthermore, after displaying the question-and-answer results or generated reply statements to the user, upon receiving the user's interactive operation, the corresponding question-and-answer data is updated based on the interactive operation through the user feedback optimization module, and the relevant data in the reference database is updated.

[0165] Furthermore, if no questions similar to the query statement are retrieved in the reference database, the query statement is marked as a task to be batch run and sent to the batch run module.

[0166] Furthermore, after updating the relevant data in the reference database, the reference database is screened, and the question statements corresponding to the answer statements with low confidence in the database are marked as tasks to be batched and sent to the batching module.

[0167] Furthermore, when the conditions for executing the batch task are met, the batch module is used to batch request the large language model, execute the tasks to be batched, generate new reply sentences corresponding to the question sentences, and store the new question-answer pairs generated by the batch module into the database.

[0168] Therefore, the overall implementation logic of the processing method proposed in the embodiment of the present disclosure can not only effectively avoid the large model from repeatedly generating answers to the same or similar questions, causing waste of resources, but also improve the richness of the answers, improve processing efficiency and reliability, and thus improve user experience.

[0169] In order to implement the above embodiment, the present disclosure further provides a processing device.

[0170] Figure 8 A schematic diagram of the structure of a processing device provided in one embodiment of the present disclosure.

[0171] like Figure 8 As shown, the processing device 800 includes: a receiving module 801 , an acquiring module 802 , a generating module 803 , a display module 804 , and an updating module 805 .

[0172] Receiving module 801, for receiving a query statement;

[0173] An acquisition module 802 is configured to acquire a reference question-answer pair associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, each of which includes a question statement and a generated first answer statement;

[0174] A generating module 803 is configured to generate a search result display page based on the reference question-answer pair, wherein the length of the second answer statement in the target question-answer pair on the search result display page is less than the length of the corresponding first answer statement;

[0175] Display module 804, configured to display an interactive interface associated with any second reply statement upon receiving a selection operation for any second reply statement, wherein the interactive interface is an interface for generating the corresponding first reply statement;

[0176] The updating module 805 is configured to update the question-answer pair data associated with any second reply in the reference database based on the interactive operation received in the interactive interface.

[0177] Optionally, the generating module 803 is specifically used for any of the following:

[0178] Generating a second answer statement in a search result display page based on the summary information corresponding to the first answer statement in the reference question-answer pair;

[0179] A second answer statement in a search result display page is generated based on a portion of a first answer statement in a reference question-answer pair.

[0180] Optionally, the search result display page further includes at least one of the following:

[0181] The matching degree between the question sentence and the query sentence in the target question-answer pair;

[0182] The confidence of the first answer sentence corresponding to the target question-answer pair;

[0183] Generation time information of the target question-answer pair;

[0184] Generate model information for the corresponding first reply sentence.

[0185] Optionally, the processing device 800 further includes:

[0186] The determination module (not shown in the figure) is used to determine the display order of the target question and answer pairs in the search result display page based on the matching degree between the question statement and the query statement in the reference question and answer pair, and the confidence of the corresponding first reply statement.

[0187] Optionally, the updating module 805 is specifically configured to perform at least one of the following:

[0188] In a case where the interactive operation is a feedback operation for the corresponding first reply sentence, updating the confidence of the corresponding first reply sentence based on the type and / or emotional direction of the feedback operation;

[0189] In the case where the interactive operation is an operation directed to an associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation.

[0190] Optionally, the processing device 800 further includes:

[0191] A first marking module (not shown in the figure) is configured to mark the first question statement corresponding to any first reply statement as a task to be run if the confidence level corresponding to any first reply statement is less than a threshold, or if the weight of the reference information source associated with any first reply statement is updated;

[0192] The first execution module (not shown in the figure) is used to execute the task to be run in batches when the execution conditions of the batch tasks are met, so as to generate a new answer statement corresponding to the first question statement.

[0193] Optionally, the batch task execution condition includes at least one of the following:

[0194] The remaining computing power resources in the system are greater than the resource threshold;

[0195] Preset time period;

[0196] The number of tasks to be run in a batch exceeds the threshold.

[0197] Optionally, the processing device 800 further includes:

[0198] A second marking module (not shown in the figure) is used to mark the query statement as a task to be run if the reference database does not contain a reference question-answer pair associated with the query statement;

[0199] The second execution module (not shown in the figure) is used to execute the task to be run in batches when the execution conditions of the task are met, so as to generate a reply statement corresponding to the query statement.

[0200] It should be noted that the explanation of the aforementioned processing method embodiment is also applicable to the processing device of this embodiment, so it will not be repeated here.

[0201] In the disclosed embodiment, a query statement is first received, and then a reference question-answer pair associated with the query statement is obtained from a reference database, and a search result display page is generated based on the reference question-answer pair. After that, when a selection operation is received for any second reply statement, an interactive interface associated with any second reply statement is displayed. Finally, based on the interactive operation received in the interactive interface, the question-answer pair data associated with any second reply in the reference database is updated. Thus, by obtaining the associated question-answer pair data from the reference database based on the query statement, and generating a search result display page based on the question-answer pair data, the search engine can be effectively prevented from repeatedly calling a large model to generate answers for the same or similar questions, thereby shortening the processing time, improving the utilization rate of computing resources and processing efficiency, and improving the richness of search results. When a selection operation is received for any question-answer pair, an interactive interface associated with the corresponding question-answer pair is displayed, and based on the interactive operation of the user in the interactive interface, the corresponding question-answer pair data in the reference database is updated, so that the data in the database can be updated in a timely manner, the reliability and accuracy of the search results can be improved, and the user experience can be enhanced.

[0202] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0203] Figure 9 A schematic block diagram of an example electronic device 900 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0204] like Figure 9As shown, the device 900 includes a computing unit 901, which can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 902 or a computer program loaded from a storage unit 908 into a RAM (Random Access Memory) 903. Various programs and data required for the operation of the device 900 can also be stored in the RAM 903. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An I / O (Input / Output) interface 905 is also connected to the bus 904.

[0205] Various components in the device 900 are connected to the I / O interface 905, including an input unit 906, such as a keyboard, a mouse, etc.; an output unit 907, such as various types of displays, speakers, etc.; a storage unit 908, such as a magnetic disk, an optical disk, etc.; and a communication unit 909, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 909 allows the device 900 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0206] The computing unit 901 can be various general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), various specialized AI (Artificial Intelligence) computing chips, various computing units that run machine learning model algorithms, a DSP (Digital Signal Processor), and any appropriate processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as the processing method. For example, in some embodiments, the processing method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 908. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 900 via the ROM 902 and / or the communication unit 909. When the computer program is loaded into the RAM 903 and executed by the computing unit 901, one or more steps of the processing method described above can be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to execute the processing method in any other appropriate manner (eg, by means of firmware).

[0207] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System on Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0208] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0209] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or apparatus. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, RAM, ROM, EPROM (Electrically Programmable Read-Only-Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0210] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0211] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: LAN (Local Area Network), WAN (Wide Area Network), the Internet, and blockchain networks.

[0212] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. The client-server relationship is established by computer programs running on the respective computers and establishing a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service ecosystem that addresses the management difficulties and poor scalability of traditional physical hosts and VPS services. The server may also be a server in a distributed system or a server integrated with blockchain.

[0213] According to an embodiment of the present disclosure, the present disclosure further provides a computer program product. When an instruction processor in the computer program product executes, the processing method proposed in the above embodiment of the present disclosure is executed.

[0214] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0215] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A processing method comprising: Receive query statements; Acquire a reference question-answer pair associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first answer statement; generating a search result display page based on the reference question-answer pair, wherein the length of the second answer sentence in the target question-answer pair on the search result display page is less than the length of the corresponding first answer sentence; When a selection operation is received for any second reply statement, an interactive interface associated with the second reply statement is displayed, wherein the interactive interface is an interface for generating the corresponding first reply statement; Based on the interactive operation received in the interactive interface, the question-answer pair data associated with any second answer in the reference database is updated.

2. The method according to claim 1, wherein Generating a search result display page based on the reference question-answer pair includes any of the following: generating a second answer statement in the search result display page based on the summary information corresponding to the first answer statement in the reference question and answer pair; A second reply statement in the search result display page is generated based on a portion of the first reply statement in the reference question and answer pair.

3. The method according to claim 1, wherein The search result display page also includes at least one of the following: The matching degree between the question statement in the target question-answer pair and the query statement; The confidence of the first reply sentence corresponding to the target question-answer pair; Generation time information of the target question-answer pair; Generate model information of the corresponding first reply statement.

4. The method according to claim 1, wherein The method further comprises: The display order of the target question-answer pairs in the search result display page is determined based on the matching degree between the question statement in the reference question-answer pair and the query statement, and the confidence of the corresponding first reply statement.

5. The method according to any one of claims 1 to 4, wherein: The updating of the question-answer pair data associated with any second answer in the reference database based on the interactive operation received on the interactive interface includes at least one of the following: In a case where the interactive operation is a feedback operation for the corresponding first reply sentence, updating the confidence of the corresponding first reply sentence based on the type and / or emotional direction of the feedback operation; In the case where the interactive operation is an operation directed to an associated reference information source, the weight of the information source associated with the corresponding first reply statement is updated based on the type, duration and / or emotional direction of the operation.

6. The method according to claim 5, wherein: The method further comprises: When the confidence level corresponding to any first reply statement is less than a threshold, or when the weight of the reference information source associated with any first reply statement is updated, the first question statement corresponding to any first reply statement is marked as a task to be run; When the execution conditions of the batch task are met, the task to be batch-run is executed to generate a new answer statement corresponding to the first question statement.

7. The method according to claim 6, wherein: The batch task execution condition includes at least one of the following: The remaining computing power resources in the system are greater than the resource threshold; Preset time period; The number of tasks to be run in a batch exceeds the threshold.

8. The method of claim 6, wherein: The method further comprises: If the reference database does not contain a reference question-answer pair associated with the query statement, marking the query statement as a task to be run; When the execution conditions of the batch task are met, the task to be batch-run is executed to generate a reply statement corresponding to the query statement.

9. A processing device comprising: A receiving module, used for receiving a query statement; an acquisition module, configured to acquire a reference question-answer pair associated with the query statement from a reference database, wherein the reference database includes a plurality of question-answer pair data, and the question-answer pair data includes a question statement and a generated first answer statement; a generating module for generating a search result display page based on the reference question-answer pair, wherein the length of the second answer statement in the target question-answer pair on the search result display page is less than the length of the corresponding first answer statement; A display module, configured to, upon receiving a selection operation for any second reply statement, display an interactive interface associated with the second reply statement, wherein the interactive interface is an interface for generating the corresponding first reply statement; An updating module is used to update the question-answer pair data associated with any second reply in the reference database based on the interactive operation received in the interactive interface.

10. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 8.

11. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-8.

12. A computer program product comprising a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

Images