Data search method and device based on AI language model

Abstract

The invention discloses a data search method and device based on an AI language model, and relates to the technical field of information retrieval. The problems that search information input by a user cannot be accurately recognized, and meanwhile accurate search results cannot be rapidly provided for the user according to the recognized search information are solved. The method specifically comprises the following steps: receiving search information input by a user, and analyzing and reconstructing the information input by the user through an AI language model; and carrying out classified retrieval on the search information through a database. The AI language model is used for analyzing the search information input by the user, keyword recognition can be quickly performed on the information input by the user, the technical field corresponding to the search information can be quickly clarified, meanwhile, the search statement is reconstructed through the keyword analyzed by the AI language model, and the search efficiency is improved. By means of the method, the search statements which are not standard enough and input by the user can become professional and accurate, information wanted by the user can be retrieved more accurately and rapidly, and the data search efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information retrieval, and in particular, to a data search method and device based on an AI language model. Background Art

[0002] With the progress of technology and social development, current search engines need to process a large amount of data and information, including articles, videos, audios, pictures, etc. in different fields. This results in a very large number of search results when the search engine executes the search process. As a result, users cannot quickly identify which search results are relevant, and usually need to view multiple search results before they can find the results they need, making it difficult for users to obtain satisfactory search results accurately and quickly.

[0003] After retrieval, a patent with the Chinese patent application number CN202310533693.x discloses a data search method, device, medium and equipment, including receiving an input search text; searching for a number of initial search results including pictures and fields that match the search text. However, since the above technical solution does not set up a corresponding analysis method that can quickly analyze and accurately identify the user input search information, there is still a problem that it cannot accurately identify the user input search information and at the same time quickly provide accurate search results to the user according to the identified search information. Summary of the Invention

[0004] The purpose of the present invention is to solve the deficiencies in the prior art, and to propose a data search method and device based on an AI language model.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A data search method based on an AI language model includes the following steps:

[0007] S1: Receive the search information input by the user, and analyze and reconstruct the user input information through the AI language model;

[0008] S2: Classify and retrieve the search information through the database;

[0009] S3: Analyze and process the retrieved data through the AI language model;

[0010] S4: Generate a retrieval result according to the analysis of the search result by the AI language model;

[0011] S5: Send the generated retrieval result to the user in the form of a retrieval report.

[0012] Preferably: in S1, the AI language model identifies the keywords in the input search information, determines the technical field to which the user input search information belongs, and then the AI language model generates new retrieval information based on the technical field and the identified keywords to complete the analysis and reconstruction of the retrieval information.

[0013] Furthermore: in S1, the search information after being analyzed and reconstructed by the AI language model is sent to the user for confirmation. After reading, the user can choose to retrieve the analyzed and reconstructed retrieval information according to needs, or can also choose to continue to use the original input search information for retrieval;

[0014] In S2, the data retrieved from the database through the search information is divided into pure document data, graphic and text data, picture data, and audio data.

[0015] More preferably: in S2, the database will identify and classify the data stored internally. The stored pure document data is stored in the pure document database, the graphic and text data is stored in the graphic and text database, the picture data is stored in the picture database, and the audio data is stored in the audio database.

[0016] As a preferred embodiment of the present invention: in S3, the AI language model identifies the search information price, identifies whether there are keywords specifying pure document, graphic and text, picture, and audio data in the search information. After identifying the corresponding keywords, it will perform targeted retrieval through the corresponding database.

[0017] As a further preferred embodiment of the present invention: in S3, when the AI language model identifies that the search data type is not specified in the search information, it will sequentially retrieve the pure document database, the graphic and text database, the picture database, and the audio database.

[0018] As a further aspect of the present invention: in S4, the AI language model arranges the retrieved data according to the relevance between the retrieved data and the user input retrieval information, and inputs the arranged retrieval results into the corresponding retrieval report in sequence according to the arrangement order.

[0019] On the basis of the foregoing solution: in S4, the AI language model will divide the retrieval report into a pure document retrieval report, a graphic and text retrieval report, a picture retrieval report, an audio retrieval report, and a comprehensive retrieval report according to the specified search data type and the corresponding type of the retrieved data.

[0020] Preferably on the basis of the foregoing solution: in S5, in each type of retrieval report, there is an option to query other types of retrieval reports.

[0021] A data search device based on an AI language model, comprising a data receiving unit for receiving user input information, a database unit for storing data, a data retrieval unit for retrieving the database, a data processing unit for running the AI language model, and a data output unit for outputting search results. The data receiving unit, the database unit, the data retrieval unit, the data processing unit, and the data output unit are communicatively connected to each other.

[0022] The beneficial effects of the present invention are as follows:

[0023] 1. By using the AI language model to analyze the search information input by the user, it is possible to quickly identify keywords in the user input information, thereby quickly clarifying the technical field corresponding to the search information. At the same time, by reconstructing the search statement with the keywords analyzed by the AI language model, the less standardized search statement input by the user can be made professional and accurate, so that the information desired by the user can be retrieved more accurately and quickly, improving the data search efficiency.

[0024] 2. By dividing the database into a pure document database, a graphic and text database, a picture database, and an audio database, it is possible to effectively classify and manage the data stored in the database. Thus, according to the user's search needs, the text, audio, and picture data required by the user can be quickly retrieved. At the same time, it is convenient to manage various types of data, improving the search efficiency and facilitating the maintenance and management of the stored data in the database.

[0025] 3. By using the AI language model to organize the retrieved data and generate a corresponding retrieval report, it is convenient for the user to browse and view the retrieved data, enabling the user to find the corresponding data they need more quickly. There is no need for the user to screen through the vast amount of information retrieved by themselves. The user only needs to view the generated retrieval report to clearly browse the corresponding search content, facilitating the user to view the search results.

[0026] 4. By sending the retrieved information reconstructed by the AI language model analysis to the customer for confirmation, the customer can choose to retrieve using the retrieved information generated by the AI language model according to their needs or use the information they input themselves for searching, improving the flexibility of data search. At the same time, the user can select to present retrieval reports of different data types on the retrieval report, increasing the richness of the data retrieved by the user and helping the user better understand and read the retrieved data.

[0027] 5. By analyzing the data type of the retrieval results required by the user through the AI language model, it is possible to accurately present the required documents, pictures, or audio for the user, thereby matching the search results in the corresponding data form according to the user's needs, facilitating the user to view and read the search results, and further improving the accuracy of data search. Brief Description of the Drawings

[0028] Figure 1 It is a schematic flow chart of a data search method based on an AI language model proposed by the present invention;

[0029] Figure 2 It is a schematic communication connection diagram of a data search device based on an AI language model proposed by the present invention. Detailed Embodiments

[0030] The technical solution of the present invention will be further described in detail below in conjunction with the detailed embodiments.

[0031] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0032] Embodiment 1:

[0033] A data search method based on an AI language model, as Figure 1 shown, includes the following steps:

[0034] S1: Receive the search information input by the user, and analyze and reconstruct the user input information through the AI language model;

[0035] S2: Classify and retrieve the search information through the database;

[0036] S3: Analyze and process the retrieved data through the AI language model;

[0037] S4: Generate a retrieval result according to the analysis of the AI language model on the search result;

[0038] S5: Send the generated retrieval result to the user in the form of a retrieval report.

[0039] In the said S1, the keywords in the input search information are identified through the AI language model to determine the technical field to which the user input search information belongs. Subsequently, the AI language model generates new retrieval information according to this technical field through the identified keywords to complete the analysis and reconstruction of the retrieval information;

[0040] In the said S1: The search information after being analyzed and reconstructed by the AI language model is sent to the user for confirmation. After reading, the user can choose to retrieve the analyzed and reconstructed retrieval information according to needs, or can also choose to continue to use the original input search information for retrieval;

[0041] In S2, the data retrieved from the database through the search information is divided into pure document data, text and image data, image data, and audio data;

[0042] In S2, the database will identify and classify the data stored internally. The stored pure document data is stored in the pure document database, the text and image data is stored in the text and image database, the image data is stored in the image database, and the audio data is stored in the audio database;

[0043] In S3, the AI language model identifies the search information price, identifies whether there are keywords specifying pure document, text and image, image, and audio data in the search information. After identifying the corresponding keywords, it will perform targeted retrieval through the corresponding database;

[0044] In S3, when the AI language model identifies that the search data type is not specified in the search information, it will sequentially retrieve the pure document database, the text and image database, the image database, and the audio database;

[0045] In S4, the AI language model arranges the retrieved data according to the relevance between the retrieved data and the user input search information, and inputs the arranged retrieval results into the corresponding retrieval report in the arranged order;

[0046] In S4, the AI language model will divide the retrieval report into a pure document retrieval report, a text and image retrieval report, an image retrieval report, an audio retrieval report, and a comprehensive retrieval report according to the specified search data type and the corresponding type of the retrieved data;

[0047] In S5, in each type of retrieval report, there is an option to query other types of retrieval reports.

[0048] By using the AI language model to analyze the search information input by the user, it is possible to quickly identify the keywords of the user input information, thereby quickly clarifying the technical field corresponding to the search information. At the same time, by reconstructing the search statement with the keywords analyzed by the AI language model, the non-standard search statement input by the user can be made professional and accurate, so as to be able to retrieve the information the user wants more accurately and quickly, improving the data search efficiency.

[0049] The AI language model used in this embodiment is a mature language model such as deepseek and ChatGPT in the prior art;

[0050] By dividing the database into a pure document database, a text and image database, a picture database, and an audio database, the data stored in the database can be effectively classified and managed. Thus, according to the user's search needs, the text, audio, and picture data required by the user can be quickly retrieved. At the same time, it is convenient to manage various types of data, improving the search efficiency and facilitating the maintenance and management of the stored data in the database.

[0051] By using an AI language model to sort out the retrieved data and generate a corresponding retrieval report, it is convenient for users to browse and view the retrieved data, enabling users to find the corresponding data they need more quickly without having to screen through the vast amount of information retrieved by themselves. Users only need to view the generated retrieval report to clearly browse the corresponding search content, facilitating the viewing of search results.

[0052] By analyzing the data type of the retrieval results required by the user through the AI language model, the required documents, pictures, or audio can be accurately presented to the user, thus matching the search results in the corresponding data form according to the user's needs, facilitating the reading and viewing of search results by users, and further improving the accuracy of data search.

[0053] By sending the retrieved information reconstructed by the AI language model analysis to the customer for confirmation, the customer can choose to use the retrieval information generated by the AI language model for retrieval or use the information they input themselves for searching, improving the flexibility of data search. At the same time, users can select retrieval reports presenting different data types on the retrieval report, increasing the richness of the retrieved data, and helping users better understand and read the retrieved data.

[0054] Application Example 1:

[0055] The user inputs "Beijing scenic spot recommendations". The AI language model analyzes the input content and divides it into two parts: "Beijing scenic spots" and "recommendations". Through the analysis of "Beijing scenic spots", it is concluded that the core of this search content is "scenic spots", so the data content mainly consists of picture data, audio data, and text and image data. Through the analysis of "recommendations", it is determined that an introduction and analysis of "scenic spots" are required, so it is determined that the data content type preferentially retrieves audio data and text and image data;

[0056] Subsequently, the text and image data related to "Beijing scenic spot recommendations" retrieved are processed by the AI language model and presented on the retrieval report, and the links to the audio data related to "Beijing scenic spot recommendations" retrieved are inserted into the retrieval report, thus generating a retrieval report related to "Beijing scenic spot recommendations" containing text and image information as well as audio information, enabling users to intuitively understand the situation of various scenic spots in Beijing in the form of text and image reading and video playback.

[0057] Application Example 2:

[0058] The user inputs "Chinese composition model essays", and the AI language model analyzes the input content, dividing it into two parts: "Chinese composition" and "model essays". Through the analysis of "Chinese composition", it is concluded that the core of this search content is "composition", so the data content mainly consists of pure document data. Through the analysis of "model essays", it is concluded that the composition topic and content need to be presented. Therefore, it is determined that the data content type preferentially retrieves pure document data;

[0059] Subsequently, the AI language model generates five options for the user to choose from based on the two keywords "Chinese composition" and "model essays", namely "Primary school Chinese composition model essays", "Junior high school Chinese composition model essays", "Senior high school Chinese composition model essays", "College Chinese composition model essays", and "Directly conduct a search". The user can select multiple options to conduct a search simultaneously. For example, the user can select "Primary school Chinese composition model essays" and "Junior high school Chinese composition model essays" to conduct a search;

[0060] Subsequently, the pure document data related to "Primary school Chinese composition model essays" and "Junior high school Chinese composition model essays" retrieved is then presented in the search report, thereby generating a relevant search report containing "Primary school Chinese composition model essays" and "Junior high school Chinese composition model essays", enabling the user to read the corresponding composition model essay search report suitable for themselves.

[0061] Application Example 3:

[0062] The user inputs "How to apply for a passport", and the AI language model analyzes the input content, dividing it into two parts: "How to apply" and "passport". Through the analysis of "How to apply", it is concluded that the core of this search content is "How to handle a business", so the data content mainly consists of picture data, audio data, and picture-text data. Through the analysis of "passport", it is concluded that the business to be handled is "passport application". Therefore, it is determined that the data content type preferentially retrieves picture-text data and audio data;

[0063] Subsequently, the picture-text data related to "Passport application process" retrieved is presented in the search report after being processed by the AI language model, and the link to the audio data related to "Applying for a passport" retrieved is inserted into the search report, generating a relevant search report on "How to apply for a passport" containing picture-text information and audio information, enabling the user to intuitively understand the passport application process and precautions in the form of picture-text reading and video playback.

[0064] Example 2:

[0065] A data search device based on an AI language model, such as Figure 2As shown in the figure, the following improvements are made in this embodiment based on Embodiment 1: It includes a data receiving unit for receiving user input information, a database unit for storing data, a data retrieval unit for retrieving the database, a data processing unit for running an AI language model, and a data output unit for outputting search results. The data receiving unit, the database unit, the data retrieval unit, the data processing unit, and the data output unit are communicatively connected to each other.

[0066] When this embodiment is in use, the database unit stores data through the corresponding data storage device. The user inputs the corresponding retrieval information through the data receiving unit. Subsequently, the data processing unit analyzes and processes the user input retrieval information through the AI language model. Then, according to the data obtained from the analysis, the data retrieval unit retrieves the database. Subsequently, the retrieved data is analyzed and processed through the AI language model, and a corresponding retrieval report is generated. Then, the retrieval report is sent to the user through the data output unit.

[0067] As mentioned above, the above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A data search method based on an AI language model, characterized in that, It includes the following steps: S1: Receive the search information input by the user, and analyze and reconstruct the user input information through the AI language model; S2: Classify and retrieve the search information through the database; S3: Analyze and process the retrieved data through the AI language model; S4: Generate a retrieval result based on the analysis of the search result by the AI language model; S5: Send the generated retrieval result to the user in the form of a retrieval report.

2. The data search method based on an AI language model according to claim 1, wherein In S1, the keywords in the input search information are identified through the AI language model to determine the technical field to which the user input search information belongs. Subsequently, the AI language model generates new retrieval information based on the identified keywords according to this technical field, completing the analysis and reconstruction of the retrieval information.

3. The data search method based on an AI language model according to claim 2, wherein In S1: The search information after being analyzed and reconstructed by the AI language model is sent to the user for confirmation. After reading, the user can choose to retrieve the analyzed and reconstructed retrieval information according to needs, or can also choose to continue to use the original input search information for retrieval; In S2, the data retrieved from the database through the search information is divided into pure document data, graphic and text data, picture data, and audio data.

4. A data search method based on an AI language model according to claim 3, characterized in that, In S2, the database will identify and classify the data stored internally. The stored pure document data is stored in the pure document database, the graphic and text data is stored in the graphic and text database, the picture data is stored in the picture database, and the audio data is stored in the audio database.

5. A data search method based on an AI language model according to claim 4, characterized in that, In S3, the AI language model identifies the search information price, identifies whether there are keywords specifying pure document, graphic and text, picture, and audio data in the search information. After identifying the corresponding keywords, it will conduct targeted retrieval through the corresponding database.

6. The data search method based on an AI language model according to claim 5, wherein In S3, when the AI language model identifies that the search data type is not specified in the search information, it will sequentially retrieve the pure document database, graphic and text database, picture database, and audio database.

7. A data search method based on an AI language model according to claim 6, characterized in that, In S4, the AI language model arranges the retrieved corresponding data according to the relevance between the retrieved data and the user input retrieval information, and inputs the arranged retrieval results into the corresponding retrieval report in sequence according to the arrangement order.

8. The data search method based on an AI language model according to claim 7, wherein, In S4, the AI language model will divide the retrieval report into a pure document retrieval report, a graphic and text retrieval report, a picture retrieval report, an audio retrieval report, and a comprehensive retrieval report according to the specified search data type and the corresponding type of the retrieved data.

9. A data search method based on an AI language model according to claim 8, characterized in that, In S5, in each type of retrieval report, there is an option to query other types of retrieval reports.

10. A data search device based on an AI language model, comprising a data receiving unit for receiving user input information, a database unit for storing data, a data retrieval unit for retrieving the database, a data processing unit for running the AI language model, and a data output unit for outputting search results, characterized in that, The data receiving unit, database unit, data retrieval unit, data processing unit, and data output unit are communicatively connected to each other.

Images