Information display method and device based on large model, equipment, medium and product

Abstract

The invention provides an information display method and device based on a large model, equipment, a medium and a product, and relates to the technical field of data processing, in particular to the technical field of artificial intelligence such as large models, natural language processing and deep learning. The information display method based on the large model comprises the steps that in response to a received query problem, matching is conducted on a field value set of a target data table corresponding to the query problem according to the query problem, at least one target field value set is obtained, the target field value set comprises multiple target field values related in semantics, and the target field value set comprises multiple target field values related in semantics; the plurality of target field values are used for reducing the semantic deviation of understanding the query problem by the large model; according to prompt information obtained based on the query problem, the at least one target field value group and the description information of the target data table, calling the large model to generate a target query statement; and displaying a query result obtained by executing the target query statement.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technologies, and particularly to artificial intelligence technology fields such as large models, natural language processing, and deep learning. More specifically, it relates to a method, apparatus, device, medium, and product for information display based on a large model. Background Art

[0002] With the advent of the big data era, the demand for data query and analysis has been increasing day by day. Since the data query method based on structured query statements has relatively high requirements for users' professional knowledge, in order to lower the threshold of data query, the data query method based on natural language has emerged as the times require. The data query method based on natural language allows users to pose query questions in natural language and automatically converts them into query statements to achieve data query and analysis. Summary of the Invention

[0003] The present disclosure provides a method, apparatus, device, medium, and product for information display based on a large model.

[0004] According to one aspect of the present disclosure, there is provided a method for information display based on a large model, including: in response to receiving a query question, matching a set of field values of a target data table corresponding to the query question according to the query question to obtain at least one target field value group, where the target field value group includes a plurality of target field values that are semantically associated, and the plurality of target field values are used to reduce the semantic deviation of the large model's understanding of the query question; calling the large model according to the prompt information obtained based on the query question, the at least one target field value group, and the description information of the target data table to generate a target query statement; and displaying the query result obtained by executing the target query statement.

[0005] According to another aspect of the present disclosure, there is provided an information display device based on a large model, including: a matching module, configured to match a set of field values of a target data table corresponding to a query question according to the query question in response to receiving the query question to obtain at least one target field value group, where the target field value group includes a plurality of target field values that are semantically associated, and the plurality of target field values are used to reduce the semantic deviation of the large model's understanding of the query question; a generation module, configured to call the large model according to the prompt information obtained based on the query question, the at least one target field value group, and the description information of the target data table to generate a target query statement; and a display module, configured to display the query result obtained by executing the target query statement.

[0006] According to another aspect of the present disclosure, there is provided an intelligent agent of artificial intelligence, configured to execute the steps of the above method.

[0007] According to another aspect of the present disclosure, there is provided an electronic device, including: one or more processors; a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0008] According to another aspect of the present disclosure, there is provided a computer-readable storage medium, on which computer programs or instructions are stored, and when the computer programs or instructions are executed by a processor, the steps of the above method are implemented.

[0009] According to another aspect of the present disclosure, there is provided a computer program product, including computer programs or instructions, and when the computer programs or instructions are executed by a processor, the steps of the above method are implemented.

[0010] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used 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

[0011] Through the following description of the embodiments of the present disclosure with reference to the drawings, the above and other objects, features and advantages of the present disclosure will become clearer. In the drawings:

[0012] Figure 1 Schematically shows a system architecture to which the information display method based on a large model can be applied according to an embodiment of the present disclosure;

[0013] Figure 2 Schematically shows a flowchart of the information display method based on a large model according to an embodiment of the present disclosure;

[0014] Figure 3 Schematically shows an example diagram of the construction process of a set of field values corresponding to a candidate data table according to an embodiment of the present disclosure;

[0015] Figure 4 Schematically shows an example diagram of the process of matching a set of field values of a target data table corresponding to a query problem to obtain at least one target field value group according to an embodiment of the present disclosure;

[0016] Figure 5 Schematically shows an example diagram of the process of calling a large model to generate a target query statement according to prompt information obtained based on a query problem, at least one target field value group, and description information of a target data table according to an embodiment of the present disclosure;

[0017] Figure 6 Schematically shows a block diagram of an information display device based on a large model according to an embodiment of the present disclosure;

[0018] Figure 7 Schematically shows a structural block diagram of an agent of a large model according to an embodiment of the present disclosure; and

[0019] Figure 8 Schematically shows a block diagram of an electronic device suitable for implementing an information display method based on a large model according to an embodiment of the present disclosure. Detailed implementation manners

[0020] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the following detailed description, for the sake of explanation, many specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments can also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily confusing the concepts of the present disclosure.

[0021] The terms used herein are merely for describing specific embodiments and are not intended to limit the present disclosure. The terms "including", "comprising", etc. used herein indicate the presence of the described features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0022] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0023] In the case of using expressions such as "at least one of A, B, and C, etc.", generally, it should be interpreted according to the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include, but is not limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.).

[0024] Based on the natural language data query method, it is necessary to convert the user's natural language query into an executable query statement through a large model.

[0025] In this process, since some field values in the natural language query input by the user may be different from the field values in the database, such as synonyms, abbreviations, spelling mistakes, etc., it is impossible to effectively process the connection between the corresponding content in the natural language and the field values in the database, resulting in the generated query statement being unable to accurately retrieve relevant information, affecting the accuracy of data query and analysis and the user experience.

[0026] To this end, embodiments of the present disclosure propose an information display solution based on a large model. For example, in response to receiving a query question, according to the query question, match a set of field values of a target data table corresponding to the query question to obtain at least one target field value group, where the target field value group includes multiple target field values that are semantically associated, and the multiple target field values are used to reduce the semantic deviation of the large model's understanding of the query question; according to the prompt information obtained based on the query question, at least one target field value group, and the description information of the target data table, call the large model to generate a target query statement; and display the query result obtained by executing the target query statement.

[0027] According to an embodiment of the present disclosure, by accurately matching the target field value group according to the query question, since the target field value group includes multiple target field values that are semantically associated, these multiple target field values can be used to reduce the semantic deviation of the large model's understanding of the query question, thereby improving the accuracy and reliability in processing natural language query questions, being able to generate accurate query statements, so as to quickly and accurately obtain and display the query results required by the user, enhancing the user experience, and improving the efficiency and quality of data query.

[0028] In the technical solution of the present invention, the processing of collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information involved all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0029] In the technical solution of the present invention, before obtaining or collecting the user's personal information, the user's authorization or consent is obtained.

[0030] Figure 1 Schematically shows a system architecture to which the information display method based on a large model according to an embodiment of the present disclosure can be applied. It should be noted that Figure 1 What is shown is only an example of a system architecture to which embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0031] As Figure 1 shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0032] Users can use at least one of the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (for example only).

[0033] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with a display screen and supporting web browsing, including but not limited to smartphones, tablets, laptop computers, desktop computers, and so on.

[0034] The server 105 can be a server that provides various services, such as a background management server that supports the websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (for example only). The background management server can analyze and process data such as received user requests, and feedback the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0035] It should be noted that the information display method based on the large model provided by the embodiments of the present disclosure can generally be executed by the server 105. Correspondingly, the information display device based on the large model provided by the embodiments of the present disclosure can generally be set in the server 105. The information display method based on the large model provided by the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105. Correspondingly, the information display device based on the large model provided by the embodiments of the present disclosure can also be set in a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105.

[0036] Alternatively, the information display method based on the large model provided by the embodiments of the present disclosure can also be executed by the first terminal device 101, the second terminal device 102, or the third terminal device 103, or can also be executed by other terminal devices different from the first terminal device 101, the second terminal device 102, or the third terminal device 103. Correspondingly, the information display device based on the large model provided by the embodiments of the present disclosure can also be set in the first terminal device 101, the second terminal device 102, or the third terminal device 103, or set in other terminal devices different from the first terminal device 101, the second terminal device 102, or the third terminal device 103.

[0037] It should be understood that Figure 1 the numbers of the terminal devices, networks, and servers in [[ ]] are merely illustrative. According to the implementation requirements, there can be any number of terminal devices, networks, and servers.

[0038] It should be noted that the sequence numbers of the respective operations in the following methods are only used as representations of the operations for description purposes and should not be regarded as indicating the execution order of the respective operations. Unless explicitly stated, the method does not need to be executed exactly in the order shown.

[0039] The system architecture to which the information display method based on a large model provided by the present disclosure can be applied has been described above. Below, taking [[ ]] Figure 2 as an example, the information display process based on a large model of the present disclosure will be further described.

[0040] Figure 2 FIG. schematically shows a flowchart of an information display method based on a large model according to an embodiment of the present disclosure.

[0041] As [[ ]] Figure 2 shown, the information display method 200 based on a large model includes operations S210 to S230.

[0042] In operation S210, in response to receiving a query question, according to the query question, a set of field values of a target data table corresponding to the query question is matched to obtain at least one target field value group, where the target field value group includes a plurality of target field values that are semantically associated, and the plurality of target field values are used to reduce the semantic deviation of the large model's understanding of the query question.

[0043] In operation S220, according to the prompt information obtained based on the query question, at least one target field value group, and the description information of the target data table, a large model is called to generate a target query statement.

[0044] In operation S230, the query result obtained by executing the target query statement is displayed.

[0045] Before executing the information display method provided by the present disclosure, a set of field values of each of a plurality of candidate data tables can be pre-generated. For each candidate data table, the set of field values can include a plurality of candidate fields and a candidate field value group corresponding to each candidate field. The candidate field value group can include at least one candidate field value, and the candidate field values in the same candidate field value group are semantically associated with each other.

[0046] In an embodiment of the present disclosure, before obtaining the user's information, the user's consent or authorization can be obtained. For example, before operation S210, a request to obtain the user's information can be sent to the user. When the user consents or authorizes to obtain the user's information, operation S210 is executed.

[0047] A query question refers to the information that the user wants to query in the form of natural language. After receiving the query question, the target data table corresponding to the query question can be determined from multiple candidate data tables. The determination method of the target data table can be configured according to actual business requirements and is not limited here.

[0048] After determining the target data table, the field value set of the target data table can be matched according to the query question to obtain at least one target field value group. A target field value group can include multiple target field values, and the semantic meanings of the target field values in the same target field value group are associated with each other to reduce the semantic deviation of the large model's understanding of the query question.

[0049] The determination method of the target field value group can be configured according to actual business requirements and is not limited here. For example, a semantic embedding model can be used to convert the query question and the field value set into high-dimensional vectors respectively, and matching can be performed through methods such as cosine similarity calculation to determine the target field value group. Alternatively, a rule-based matching method can be adopted, establishing mapping rules between natural language vocabulary and each field value in the data table in advance, and performing one-by-one matching on the vocabulary in the query question according to these rules to determine the target field value group.

[0050] After obtaining at least one target field value group, a prompt message can be constructed based on the query question, at least one target field value group, and the description information of the target data table. The prompt message is used to be input into the large model to assist it in generating the target query statement. The description information of the target data table refers to the description of relevant information such as the content and structure of the target data table, such as the field categories and data ranges included in the table.

[0051] The prompt message can include the query question, multiple target field values in each target field value group, and reference query statements, etc., providing context and reference for the large model so that it can more accurately generate the target query statement that meets the user's needs. The target query statement can accurately express the user's query requirements for data and be executed in the database to obtain the data required by the user and display the query result.

[0052] The generation method of the target query statement can be configured according to actual business requirements and is not limited here. For example, the enhanced prompt words can be constructed by combining the filtered target field value groups with the query question and the description information of the target data table corresponding to the query question, and the enhanced prompt words can be input into the large model to generate the target query statement for the query question. Alternatively, according to the semantic information of the query question, a corresponding template can be designed, and the specific elements in the query question and the corresponding elements in each target field value group can be filled into the template to form prompt information, and then input into the large model to generate the target query statement.

[0053] The query results can be presented to the user in a user-friendly manner. For example, according to the type and characteristics of the query results, an appropriate chart type (such as bar chart, line chart, pie chart, etc.) can be selected to visually present the data to help the user understand and analyze the data more clearly. In addition, when presenting the query results, at least one target field value group can be presented to enhance interpretability.

[0054] After obtaining the query statement, the query statement can be executed and the query results can be presented. The presentation form of the query results can be configured according to actual business requirements and is not limited here. For example, the presentation form of the query results can include at least one of the following: table form, chart form, and text form, etc. The table form can be used to present structured query results. The chart form can include bar charts, line charts, pie charts, and scatter plots. Bar charts are used to show comparisons of different categories of data. Line charts are used to show the trend of data changes over time. Pie charts are used to show the proportion of each part to the whole. Scatter plots are used to show the relationship between two variables. The text form is used to give a summary description of the query results.

[0055] According to an embodiment of the present disclosure, by accurately matching the target field value groups according to the query question, since the target field value groups include multiple target field values that are semantically related, these multiple target field values can be used to reduce the semantic deviation of the large model's understanding of the query question, thereby improving the accuracy and reliability in processing natural language query questions, being able to generate accurate query statements, thus quickly and accurately obtaining and presenting the query results required by the user, enhancing the user experience, and improving the efficiency and quality of data query.

[0056] In one embodiment, the semantic overlap degree between multiple target field values satisfies a preset overlap degree condition. The semantic overlap degree is an index used to measure the degree of semantic overlap between multiple target field values, that is, the degree of consistency in the expression of meaning of multiple target field values. The preset overlap degree condition refers to the standard preset for judging whether the semantic overlap degree between multiple field values meets the requirements. For example, the preset overlap degree condition can be that the semantic overlap degree needs to reach above 0.8.

[0057] The determination method of semantic overlap degree can be configured according to actual business requirements and is not limited herein. For example, a semantic embedding model can be used to convert the target field values into vectors, and the cosine similarity between the vectors can be calculated to measure the semantic overlap degree. Alternatively, the semantic similarity can be manually annotated, and domain experts can subjectively judge the semantic overlap degree of multiple field values.

[0058] The target field values can include at least one of the following: synonyms, abbreviations, short names, and full names. For example, for the target field value of "net profit", the same target field value group can include the amount obtained by subtracting income tax from the total profit of the enterprise in the current period, profit, Net Profit, P, etc. "Net Profit" can be understood as the full name of "net profit", "P" can be understood as the short name or abbreviation of "net profit", and "the amount obtained by subtracting income tax from the total profit of the enterprise in the current period" can be understood as a synonym of "net profit".

[0059] According to the embodiments of the present disclosure, by ensuring that the semantic overlap degree between multiple target field values meets the preset overlap degree condition, the same or similar semantics represented by different expressions in a natural language query can be captured, so as to more accurately match the correct target field value group, enhance the understanding of user intentions, effectively improve the accuracy of data query in semantic understanding, and help improve the accuracy and reliability of query results.

[0060] In one embodiment, the field value set includes multiple candidate fields in the candidate data table and the respective candidate field value groups of each candidate field; the field value sets corresponding to the respective candidate data tables are obtained by the following method: performing structured information extraction on the candidate data table to obtain multiple candidate fields and the respective multiple candidate field values of each candidate field; and for each candidate field, respectively performing semantic feature extraction on the multiple candidate field values to obtain candidate field value groups, and using the multiple candidate fields and the respective candidate field value groups of each candidate field as the field value set.

[0061] For each candidate data table, structured information extraction refers to the process of extracting structured fields and corresponding field value information from the candidate data table. For example, the field names of the respective candidate fields and the corresponding candidate field values are extracted from the candidate data table through a database query statement. The specific method of structured information extraction can be configured according to actual business requirements and is not limited herein.

[0062] In one example, tools or functions provided by a database management system can be used to perform structured queries and extractions on the candidate data table. For example, the SELECT statement is used to extract the field names of each candidate field and the corresponding candidate field values in the candidate data table. Alternatively, data mining techniques can be employed to mine the field names of each structured candidate field and the corresponding candidate field values from the candidate data table. For example, the data in the candidate data table is analyzed through a clustering algorithm to extract the field names of each potential candidate field and the corresponding candidate field values.

[0063] For each candidate field value, semantic feature extraction refers to the process of converting the candidate field value into a form that can represent its semantic features. For example, the candidate field value is converted into a vector representation to reflect its semantic features. The specific manner of semantic feature extraction can be configured according to actual business requirements and is not limited herein.

[0064] In one example, a semantic embedding model can be utilized to perform semantic feature extraction on the candidate field value, convert it into a vector representation, and form a candidate field value group. For example, "net profit", "Net Profit", "P", and "the amount obtained by subtracting income tax from the total profit of an enterprise during the current period" are respectively converted into semantic features through the semantic embedding model to form a candidate field value group. Alternatively, the method of manual annotation and semantic encoding can be adopted to perform semantic feature extraction on each candidate field value respectively. For example, an expert performs semantic annotation on the candidate field value and then converts it into a specific semantic encoding to form a candidate field value group.

[0065] According to the embodiments of the present disclosure, by extracting and processing the candidate fields and the corresponding candidate field values in the candidate data table, a field value set containing rich semantic information can be effectively constructed, providing a basis for subsequent query understanding and semantic matching, enabling more accurate understanding of the user's query intention during the process of data query based on the query problem, improving the accuracy and relevance of the query results, and contributing to enhancing the performance and reliability in dealing with complex query tasks.

[0066] The following will utilize Figure 3 , to provide an exemplary illustration of the process of constructing a field value set corresponding to the candidate data table.

[0067] Figure 3 FIG. shows an example schematic diagram of the construction process of a field value set corresponding to a candidate data table according to an embodiment of the present disclosure.

[0068] As Figure 3As shown in the figure, in 300, for any candidate data table 310, structured information extraction can be performed on the candidate data table 310 to obtain the structured information 320 of the candidate data table 310. The structured information 320 may include a plurality of candidate fields and a plurality of candidate field values for each candidate field. For example, the plurality of candidate fields may include candidate field 1, candidate field 2,..., candidate field P. P is a positive integer.

[0069] Taking candidate field 2 330 and the corresponding plurality of candidate field values as an example, the construction process of the field value set will be described below.

[0070] Candidate field 2 330 may correspond to candidate field value 1, candidate field value 2,..., candidate field value Q. Q is a positive integer. For each candidate field value, semantic feature extraction can be performed on the plurality of candidate field values respectively, and the semantic feature obtained for each candidate field value is used as the candidate field value group 340 corresponding to candidate field 2 330. For example, the candidate field value group 340 may include the semantic feature 341 corresponding to candidate field value 1 331, the semantic feature 342 corresponding to candidate field value 2 332, and the semantic feature 34Q corresponding to candidate field value Q 33Q. On this basis, candidate field 2 330 and candidate field value group 340 can be used as a part of the field value set 350, and by analogy, each candidate field and the corresponding candidate field value group can be used as the field value set 350.

[0071] In one embodiment, operation S210 may include the following operations: performing semantic matching on the description information of each of the plurality of candidate data tables according to the query problem to determine the target data table for the query problem; and performing semantic matching on the field value set corresponding to the target data table according to the query problem to obtain at least one target field value group.

[0072] The specific manner of performing semantic matching on the query problem and the description information of each of the plurality of candidate data tables can be configured according to actual business requirements and will not be limited here. For example, a semantic embedding model can be used to convert the query problem and the description information of each candidate data table into vectors respectively, calculate the similarity between the vector of the query problem and the vectors of each description information, and determine the target data table according to the similarity. Alternatively, the coincidence degree and semantic relevance of keywords in the query problem and the description information of each of the plurality of candidate data tables can be counted respectively, and the target data table can be determined according to the matching degree obtained from the comprehensive coincidence degree and semantic relevance.

[0073] The specific method for semantic matching between the query problem and the set of field values in the target data table can be configured according to actual business requirements and is not limited herein. For example, a semantic embedding model can be used to convert the query problem and each of the multiple candidate field values of each candidate field in the set of field values into vectors respectively, calculate the similarity between the vector of the query problem and the vectors of the multiple candidate field values of each candidate field, and determine a target field value group in the target data table according to the similarity. Alternatively, the coincidence degree and semantic relevance between the query problem and each of the multiple candidate field values of each candidate field can be counted respectively, and a target field value group can be determined in the target data table according to the matching degree obtained from the comprehensive coincidence degree and semantic relevance.

[0074] According to an embodiment of the present disclosure, by using semantic matching technology, a target data table is screened out from multiple candidate data tables, and a suitable target field value group is further matched from the target data table, which can quickly and accurately determine the target data table and the target field value group most relevant to the query problem. This not only improves the query efficiency, reduces unnecessary data search and processing time, but also enhances the accuracy of the query and ensures the consistency between the query result and the user's needs.

[0075] In one embodiment, according to the query problem, semantic matching is performed on the set of field values corresponding to the target data table to obtain at least one target field value group, which may include the following operations: for each candidate field, performing semantic matching between the query problem and the candidate field value group to obtain the matching degree between the candidate field and the query problem; and determining the candidate field value group of the candidate field corresponding to the top N matching degrees among the sorted multiple matching degrees as the target field value group, where N is a positive integer.

[0076] The target field value set may include multiple candidate fields and the candidate field value group of each candidate field. After receiving the query problem, semantic matching can be performed between the query problem and the candidate field value group of each candidate field respectively to obtain the matching degree between the candidate field and the query problem. The matching degree is an index for measuring the semantic similarity degree between the candidate field and the query problem. For example, the matching degree may be between 0 and 1, and the higher the value of the matching degree, the more similar the candidate field is to the query problem.

[0077] In one embodiment, the candidate field value group includes the candidate semantic features of multiple candidate field values; for each candidate field, performing semantic matching between the query problem and the candidate field value group to obtain the matching degree between the candidate field and the query problem may include the following operations: respectively determining the similarity between the semantic feature of the query problem and each candidate semantic feature to obtain multiple similarities; and determining the matching degree according to the multiple similarities.

[0078] Semantic features refer to the abstract representation of the core meaning and key information expressed by natural language texts, which can reflect the semantic content and focus of the texts. For example, a natural language text can be converted into a high-dimensional vector through a semantic embedding model to capture its semantic features. Alternatively, the bag-of-words model or the TF-IDF model can also be used to represent candidate field values as vectors, where each element in the vector represents the occurrence frequency or importance of a word, thereby obtaining candidate semantic features.

[0079] After obtaining the semantic features of the query question, the similarity between the semantic features of the query question and each candidate semantic feature can be determined respectively. Similarity is used to measure the degree of similarity between the semantic features of the query question and the candidate semantic features, and can be represented in numerical form. The larger the numerical value, the higher the degree of similarity.

[0080] For each candidate field, a semantic embedding model can be used to convert the query question and each candidate field value under this candidate field into high-dimensional vector representations respectively, and then calculate the similarity between the semantic features of the query question and each candidate semantic feature. After obtaining the similarities of each candidate field value under this candidate field respectively, the average value of multiple similarities can be determined as the matching degree. In one example, the determination method of the matching degree can be shown as the following formula (1).

[0081] (1);

[0082] Where, represents the query question, represents the candidate field value group, represents the matching degree, represents the semantic feature of the query question, represents the semantic feature of the candidate field value group, represents the average value of cosine similarities, represents the number of candidate field values in the candidate field value group, represents the i-th candidate field value in the candidate field value group, represents the semantic feature of the i-th candidate field value.

[0083] Alternatively, for each candidate field, after obtaining the similarities between each candidate field value under this candidate field and the query question respectively, the similarity with the largest numerical value among multiple similarities can be used as the matching degree between this candidate field and the query question.

[0084] Alternatively, for each candidate field, after obtaining the similarities between each candidate field value under this candidate field and the query question respectively, the similarity with the median numerical value among multiple similarities can also be used as the matching degree between this candidate field and the query question.

[0085] Alternatively, weights can be set in advance for each candidate field value. For each candidate field value, an intermediate similarity can be determined based on the weight and the similarity of the candidate field value. For each candidate field, the matching degree between the candidate field and the query problem can be determined based on the respective intermediate similarities of each candidate field value.

[0086] Alternatively, a method combining keyword matching and semantic analysis can be adopted, that is, first perform word segmentation on the query problem and the candidate field value to extract keywords; then calculate the matching degree of the keywords, and at the same time use the word vector model to calculate the semantic similarity between the keywords, and obtain the final similarity by integrating the keyword matching degree and the semantic similarity.

[0087] According to the embodiments of the present disclosure, by calculating the semantic similarity between the query problem and the candidate field value group and determining the matching degree, the correlation between the candidate field and the query problem can be effectively quantified, which helps to accurately screen out the fields most relevant to the query problem, improve the understanding ability of the query problem, and thus generate more accurate query results.

[0088] After obtaining the respective matching degrees of each candidate field, the candidate fields can be sorted according to the matching degrees to obtain multiple sorted candidate fields. On this basis, the first N candidate field value groups in the candidate field value groups of each of the multiple sorted candidate fields can be determined as the target field value groups. The target field value group refers to the field value group that is most semantically relevant to the query problem.

[0089] According to the embodiments of the present disclosure, through semantic matching and sorting and screening, calculating the matching degree and selecting the field value groups corresponding to the first N matching degrees can effectively determine the field value group most relevant to the query problem from multiple candidate fields, which not only enhances the semantic understanding ability of the query problem, but also improves the data query efficiency, enables users to quickly obtain the required information, and improves the user experience.

[0090] The following will use Figure 4 to exemplarily illustrate the process of determining the target field value group.

[0091] Figure 4 Schematically shows an example diagram of a process of matching a field value set of a target data table corresponding to a query problem according to an embodiment of the present disclosure to obtain at least one target field value group.

[0092] As Figure 4 shown, in 400, after receiving the query problem 410, the description information of each of the multiple candidate data tables can be semantically matched according to the query problem 410 to determine the target data table 440 for the query problem 410.

[0093] For example, multiple candidate data tables may include candidate data table 421, candidate data table 422, …, candidate data table 42S, where S is a positive integer. After receiving the query question 410, the query question 410 can be semantically matched with the description information 431 of the candidate data table 421, the description information 432 of the candidate data table 422, …, and the description information 43S of the candidate data table 42S respectively to determine the target data table 440.

[0094] After obtaining the target data table 440, a set of field values 450 of the target data table 440 can be obtained. The set of field values 450 may include multiple candidate fields and a candidate field value group for each candidate field. For example, the multiple candidate fields may include candidate field 1, candidate field 2, and candidate field 3. The candidate field value group of candidate field 1 is {vf11, vf12, vf13}, the candidate field value group of candidate field 2 is {vf21, vf22, vf23}, and the candidate field value group of candidate field 3 is {vf31, vf32, vf33}. vf11 refers to the candidate semantic feature of the candidate field value 1 of candidate field 1, and the other representations will not be elaborated here.

[0095] For each candidate field, the similarity between the semantic feature 460 of the query question 410 and each candidate semantic feature can be determined respectively to obtain multiple similarities; and based on the multiple similarities, the matching degree between the candidate field and the query question can be determined to obtain the matching degree 470 between each candidate field and the query question.

[0096] For example, for candidate field 1, the similarity between the semantic feature vq and the candidate semantic feature vf11, the similarity between the semantic feature vq and the candidate semantic feature vf12, and the similarity between the semantic feature vq and the candidate semantic feature vf13 can be determined, and based on the above three similarities, the matching degree between candidate field 1 and the query question 460 can be determined . And so on, the matching degree between candidate field 2 and the query question 460 can be determined , the matching degree between candidate field 3 and the query question 460 .

[0097] After obtaining the matching degrees of the multiple candidate fields respectively, the matching degrees can be sorted to obtain the sorted multiple matching degrees, and the candidate field value groups of the top N matching degrees among the sorted multiple matching degrees can be determined as the target field value group 480.

[0098] In one embodiment, after obtaining the set of field values for each candidate data table, the following operations may also be performed: in response to an update of the candidate data table, compare the candidate data table before the update and the candidate data table after the update to determine at least one updated field; and update the set of field values corresponding to the candidate data table according to the at least one updated field to obtain an updated set of field values.

[0099] The monitoring of whether the candidate data table is updated and the determination method of which specific updated fields are updated in the case of an update can be configured according to actual business requirements and are not limited herein. In one example, it can be implemented through a data comparison tool provided by the database management system. For example, the data comparison tool can be used to compare the candidate data tables before and after the update row by row and field by field, and record the updated fields that have changed. Alternatively, it can be implemented through the change log of the database. For example, the field update information recorded in the change log can be analyzed to determine the updated fields.

[0100] After determining the updated fields that have been updated, for the determined updated fields, new candidate field values can be extracted from the candidate data table after the update to perform a partial update on the old set of field values. In one example, if the proportion of the updated fields in all candidate fields in the candidate data table is greater than a preset threshold, it indicates that there are more updated contents in the candidate data table. In this case, the candidate data table can also be updated as a whole directly according to the new candidate data table.

[0101] According to the embodiments of the present disclosure, by timely detecting the update of the candidate data table, accurately determining the updated fields and updating the set of field values in a targeted manner, the problem of inaccurate query results caused by data update is avoided, the efficiency of data update is improved, unnecessary resource consumption is reduced, the adaptability and performance in dealing with a dynamic data environment are enhanced, and it can ensure that the data relied on by the query problem is always the latest, which helps to improve the reliability and accuracy of data query.

[0102] In one embodiment, operation S220 may include the following operations: based on the query problem and the description information, use a large model to filter the target set of field values to obtain at least one reference set of field values; based on a preset prompt template, splice the query problem, the at least one reference set of field values, and the description information of the data table to obtain a prompt message; and based on the prompt message, call the large model to generate a target query statement.

[0103] The description information refers to the information that describes the content, structure, fields, etc. of the data table. After obtaining at least one group of target field values, the recalled group of target field values can be provided to the large model to let the large model infer which group of field values should be selected. The specific screening method for the group of target field values can be configured according to the actual business requirements and is not limited here.

[0104] In one example, the semantic understanding and generation capabilities of the large model can be utilized to screen the group of target field values based on the query question and the description information. Alternatively, the method of rule matching can also be used to screen the group of target field values based on the keywords in the query question and the keywords in the description information. For example, through keyword extraction technology, keywords are extracted from the query question and the description information, and then matched with the field values in the group of target field values to screen out the qualified reference field value groups.

[0105] The preset prompt template refers to the pre-designed format and structure for splicing prompt information, which specifies how to combine the reference field value group, the query question, and the description information of the data table into the prompt information. The splicing method of the prompt information can be configured according to the actual business requirements and is not limited here.

[0106] In one example, the preset prompt template can be "Reference field value group: {ref_value}. Now there is a query question: {question}, and the description information of the corresponding data table: {table_desc}", and the specific content can be filled in according to this preset prompt template to obtain the prompt information. In another example, the method of combining template filling and natural language generation can be adopted. For example, according to the logical structure of the preset prompt template, natural language generation technology is used to organize the reference field value group, the query question, and the description information of the data table into a smooth and natural prompt text. After obtaining the prompt information, the prompt information can be used to guide the large model to generate the target query statement.

[0107] According to the embodiments of the present disclosure, by using the large model to screen the group of target field values, the semantic understanding ability is enhanced, and the prompt information is generated in combination with the preset prompt template, which can effectively guide the large model to generate accurate target query statements, improve the understanding and response ability of the question-answering system to user queries, and ensure the accuracy and relevance of the query results.

[0108] The following will use Figure 5 to exemplarily illustrate the generation process of the target query statement.

[0109] Figure 5Schematically shown is an example schematic diagram of a process of obtaining prompt information based on description information of a query problem, at least one set of target field values, and a target data table according to an embodiment of the present disclosure, and invoking a large model to generate a target query statement.

[0110] As Figure 5 shown, in 500, after obtaining a set of target field values for query problem 501, based on query problem 501 and description information 502, the large model M510 can be used to screen the set of target field values to obtain at least one set of reference field values 503.

[0111] After obtaining at least one set of reference field values 503, based on a preset prompt template 504, query problem 501, at least one set of reference field values 503, and description information 502 can be concatenated to obtain prompt information 505. On this basis, based on prompt information 505, the large model M510 can be invoked to generate a target query statement 506.

[0112] The above is only an exemplary embodiment, but not limited thereto. Other information display methods based on large models known in the art may also be included, as long as multiple target field values can be used to reduce the semantic deviation of the large model's understanding of the query problem to improve the efficiency and quality of data query.

[0113] Based on the above information display method based on a large model, the present invention also provides an information display device based on a large model. The following will be combined with Figure 6 to describe this device in detail.

[0114] Figure 6 Schematically shown is a block diagram of an information display device based on a large model according to an embodiment of the present disclosure.

[0115] As Figure 6 shown, the information display device 600 based on a large model may include a matching module 610, a generating module 620, and a display module 630.

[0116] The matching module 610 is configured to, in response to receiving a query problem, match a set of field values of a target data table corresponding to the query problem according to the query problem to obtain at least one set of target field values, where the set of target field values includes multiple target field values that are semantically associated, and the multiple target field values are used to reduce the semantic deviation of the large model's understanding of the query problem.

[0117] The generating module 620 is configured to generate a target query statement by invoking a large model according to prompt information obtained based on description information of the query problem, at least one set of target field values, and the target data table.

[0118] A display module 630 for displaying the query results obtained by executing the target query statement.

[0119] According to an embodiment of the present disclosure, the matching module 610 may include a first matching sub-module and a second matching sub-module.

[0120] The first matching sub-module is configured to semantically match the description information of each of the multiple candidate data tables according to the query question to determine the target data table for the query question.

[0121] The second matching sub-module is configured to semantically match the set of field values corresponding to the target data table according to the query question to obtain at least one target field value group.

[0122] According to an embodiment of the present disclosure, the set of field values includes multiple candidate fields in the candidate data table and the candidate field value groups corresponding to each candidate field.

[0123] According to an embodiment of the present disclosure, the set of field values corresponding to each candidate data table is obtained by: extracting structured information from the candidate data table to obtain multiple candidate fields and multiple candidate field values corresponding to each candidate field; and for each candidate field, respectively performing semantic feature extraction on the multiple candidate field values to obtain candidate field value groups, and taking the multiple candidate fields and the candidate field value groups corresponding to each candidate field as the set of field values.

[0124] According to an embodiment of the present disclosure, the second matching sub-module may include a matching unit and a determining unit.

[0125] The matching unit is configured to, for each candidate field, semantically match the query question with the candidate field value group to obtain the matching degree between the candidate field and the query question.

[0126] The determining unit is configured to determine the candidate field value groups corresponding to the candidate fields corresponding to the top N matching degrees among the sorted multiple matching degrees as the target field value groups, where N is a positive integer.

[0127] According to an embodiment of the present disclosure, the candidate field value group includes the candidate semantic features of the multiple candidate field values.

[0128] According to an embodiment of the present disclosure, for each candidate field, the matching unit may include a first determining sub-unit and a second determining sub-unit.

[0129] The first determining sub-unit is configured to respectively determine the similarity between the semantic feature of the query question and each candidate semantic feature to obtain multiple similarities.

[0130] The second determining sub-unit is configured to determine the matching degree according to the multiple similarities.

[0131] According to an embodiment of the present disclosure, the large model-based information display device 600 may further include a comparison module and an update module.

[0132] The comparison module is configured to, in response to an update of the candidate data table, compare the candidate data table before the update and the candidate data table after the update to determine at least one updated field.

[0133] The update module is configured to update the set of field values corresponding to the candidate data table according to at least one updated field to obtain an updated set of field values.

[0134] According to an embodiment of the present disclosure, the generation module 620 may include a screening sub-module, a splicing sub-module, and a generation sub-module.

[0135] The screening sub-module is configured to, based on the query problem and the description information, use the large model to screen the target set of field values to obtain at least one reference set of field values.

[0136] The splicing sub-module is configured to splice the query problem, at least one reference set of field values, and the description information based on a preset prompt template to obtain prompt information.

[0137] The generation sub-module is configured to, based on the prompt information, call the large model to generate a target query statement.

[0138] According to an embodiment of the present disclosure, the semantic overlap degrees between multiple target field values satisfy a preset overlap degree condition, and the target field values include at least one of the following: synonyms, abbreviations, short names, and full names.

[0139] Figure 7 Schematically shows a structural block diagram of an agent of a large model according to an embodiment of the present disclosure.

[0140] In an embodiment of the present disclosure, inspired by the von Neumann architecture in modern computer theory, as Figure 7 shown, the AI agent 700 may include five core modules: an input module 710, a control module 720, a storage module 730, an arithmetic module 740, and an output module 750.

[0141] The input module 710 is responsible for receiving or sensing information such as queries, requests, instructions, signals, or data from the outside world (such as users or the external environment) and converting it into a format that the AI agent 700 can understand and process. The input module 710 is the primary link for the AI agent 700 to interact with the outside world. It enables the AI agent 700 to efficiently and accurately obtain necessary "sensory" information from the outside world and respond to this information.

[0142] In the example, the input module 710 may input the query problem described above.

[0143] The control module 720 is the core support for the AI agent 700 to handle complex tasks. During the model training phase, the control module 720 can execute the information display method based on the large model described above.

[0144] In the example, during operation, the control module 720 will continuously interact with the storage module 730, the operation module 740, and / or the output module 750. However, it should be noted that in the embodiments of the present disclosure, the control module 720 acts as a single initiator to initiate communication with the storage module 730, the operation module 740, and / or the output module 750, and there is no communication coupling between the storage module 730, the operation module 740, and the output module 750.

[0145] In the example, the performance of the control module 720 can be closely related to the large model on which the AI agent 700 is based. To fully utilize the capabilities of the large model, the internal structure of the control module 720 can be designed to be highly configurable and extensible to handle various different types of tasks and requirements in real-world scenarios.

[0146] The storage module 730 can be responsible for memorizing the set of field values corresponding to each candidate data table. The set of field values corresponding to each candidate data table as described above can be included in the storage module 730.

[0147] In the example, after the AI agent 700 receives an evaluation request, the AI agent 700 can trigger the information display process based on the large model, obtain the set of field values corresponding to each candidate data table from the storage module 730, and feedback it to the control module 720. Then, the control module 720 can pass the feedback set of field values corresponding to each candidate data table to the output module 750.

[0148] The operation module 740 can be regarded as a predefined tool library. Tools for determining semantic features and tools for calculating similarity as described above can be included in the operation module 740.

[0149] In the example, when the AI agent 700 needs to process data, relevant tools can be called from the operation module 740 and feedback to the control module 720. Then, the control module 720 can use the feedback tools to process the query problem and obtain the query result. It can be understood that although the large model has excellent language understanding and generation capabilities, like humans, the tasks it can solve without any tools are very limited. When the AI agent 700 is given the ability to call tools, it can achieve tasks such as determining semantic features with the help of tools for determining semantic features and calculating similarity with the help of tools for calculating similarity.

[0150] During the model training phase, the output module 750 can output the query results described above.

[0151] The AI agent 700 according to the embodiments of the present disclosure can simply and effectively improve the degree of intelligence, and improve flexibility and versatility.

[0152] Figure 8 A block diagram of an electronic device suitable for implementing an information display method based on a large model according to an embodiment of the present disclosure is schematically shown. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0153] As Figure 8 shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other through a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0154] Multiple components in the device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, an optical disc, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0155] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 executes the various methods and processes described above, such as the information display method based on a large model. For example, in some embodiments, the information display method based on a large model can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the XX method described above can be executed. Alternatively, in other embodiments, the computing unit 801 can be configured to execute the information display method based on a large model in any other suitable manner (e.g., by means of firmware).

[0156] Various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), systems on a chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a dedicated or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0157] The program code for implementing the methods 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, such that when the program code is executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0158] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0159] For purposes of providing an 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 a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, 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, speech, or tactile input).

[0160] The systems and techniques described herein can be implemented in a computing system that includes back-end 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 front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end 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 a communication network include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0161] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

[0162] It should be understood that the various forms of processes shown above can be used, with steps reordered, added or deleted. For example, the steps described in the present disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solution disclosed in the present disclosure can be achieved, and no limitation is imposed herein.

[0163] The above specific embodiments do not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub - combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. An information display method based on a large model, comprising: In response to receiving a query question, matching a set of field values of a target data table corresponding to the query question according to the query question to obtain at least one target field value group, wherein the target field value group includes a plurality of target field values that are semantically associated, and the plurality of target field values are used to reduce the semantic deviation of the large model's understanding of the query question; Invoking the large model according to the prompt information obtained based on the query question, the at least one target field value group, and the description information of the target data table to generate a target query statement; and Displaying the query result obtained by executing the target query statement.

2. The method according to claim 1, wherein The step of matching the set of field values of the target data table corresponding to the query question according to the query question to obtain at least one target field value group includes: Performing semantic matching on the description information of each of the plurality of candidate data tables according to the query question to determine the target data table for the query question; and Performing semantic matching on the set of field values corresponding to the target data table according to the query question to obtain the at least one target field value group.

3. The method according to claim 2, wherein, The set of field values includes a plurality of candidate fields in the candidate data tables and a candidate field value group for each of the candidate fields; The set of field values corresponding to each of the candidate data tables is obtained by the following method: Performing structured information extraction on the candidate data table to obtain the plurality of candidate fields and a plurality of candidate field values for each of the candidate fields; and And For each of the candidate fields, respectively performing semantic feature extraction on the plurality of candidate field values to obtain the candidate field value group, and using the plurality of candidate fields and the candidate field value group for each of the candidate fields as the set of field values.

4. The method according to claim 2, wherein The step of performing semantic matching on the set of field values corresponding to the target data table according to the query question to obtain the at least one target field value group includes: For each of the candidate fields, performing semantic matching between the query question and the candidate field value group to obtain the matching degree between the candidate field and the query question; and Determining the candidate field value group of the candidate field corresponding to the top N matching degrees among the sorted plurality of matching degrees as the target field value group, where N is a positive integer.

5. The method according to claim 4, wherein The candidate field value group includes candidate semantic features of a plurality of candidate field values; The step of, for each of the candidate fields, performing semantic matching between the query question and the candidate field value group to obtain the matching degree between the candidate field and the query question includes: Respectively determining the similarity between the semantic feature of the query question and each of the candidate semantic features to obtain a plurality of similarities; and And Determining the matching degree according to the plurality of similarities.

6. The method according to claim 2, further comprising: In response to the candidate data table being updated, comparing the candidate data table before the update and the candidate data table after the update to determine at least one updated field; and And Update the set of field values corresponding to the candidate data table according to the at least one update field to obtain an updated set of field values.

7. The method according to any one of claims 1 to 6, wherein The method of invoking a large model to generate a target query statement according to the prompt information obtained based on the query problem, the at least one target field value group, and the description information of the target data table includes: Based on the query problem and the description information, use the large model to filter the target field value group to obtain at least one reference field value group; Based on a preset prompt template, splice the query problem, the at least one reference field value group, and the description information to obtain the prompt information; and Based on the prompt information, invoke the large model to generate the target query statement.

8. The method according to any one of claims 1 to 6, wherein The semantic overlap degree between the multiple target field values satisfies a preset overlap degree condition, and the target field value includes at least one of the following: synonyms, abbreviations, short names, and full names.

9. An information display device based on a large model, including: A matching module, configured to, in response to receiving a query problem, match a set of field values of a target data table corresponding to the query problem according to the query problem to obtain at least one target field value group, where the target field value group includes multiple target field values that are semantically associated, and the multiple target field values are used to reduce the semantic deviation of the large model's understanding of the query problem; A generation module, configured to invoke the large model to generate a target query statement according to the prompt information obtained based on the query problem, the at least one target field value group, and the description information of the target data table; and A display module, configured to display the query result obtained by executing the target query statement.

10. An intelligent agent of artificial intelligence, configured to execute the method according to any one of claims 1 to 8.

11. An electronic device, including: One or more processors; A memory, configured to store one or more computer programs, Characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 8.

12. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, The computer program or instruction, when executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

13. A computer program product, comprising a computer program or instructions, characterized in that, The computer program or instruction, when executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

Images