Research report generation method and device, electronic equipment and storage medium

Abstract

The invention relates to the field of natural language processing, in particular to an investigation report generation method and device, electronic equipment and a storage medium. The method comprises the following steps: in response to a keyword input by a user, acquiring data information related to the keyword from a preset data source; screening content associated with the keyword in the data information based on a preset screening loss function; and based on the content associated with the keyword, generating an investigation report associated with the keyword.

Description

Technical Field

[0001] The present disclosure relates to the field of natural language processing, and more particularly, to a method, apparatus, electronic device, and storage medium for generating a research report. Background Art

[0002] Currently, the demand for research reports in academic and business fields is increasing continuously, especially in scenarios such as enterprise technology R & D, market analysis, and strategic formulation. However, there are often significant deficiencies in the current commercial process of generating research reports, mainly reflected in the double consumption of time and cost. In order to obtain valuable information, enterprises and researchers have to invest a large amount of manpower and funds in collecting, organizing, and analyzing data.

[0003] Traditional research methods often require mobilizing multiple researchers to manually search for, screen, and read a large number of academic documents, and finally write a research report. This process is inefficient and it is difficult to ensure the comprehensiveness and consistency of the research results. Due to the differences in subjective judgments of researchers, the generated reports may miss key information or lack sufficient depth. At the same time, the quality of the research report is greatly limited by the experience and knowledge reserve of the researchers, making it difficult to achieve the expected research effect. In addition, the process of manual research involves a large amount of repetitive labor and is extremely prone to errors, especially in the case of screening and summarizing a large amount of information. Summary of the Invention

[0004] An object of the present disclosure is to provide a method for generating a research report, which solves the problems of low efficiency and insufficient accuracy of manual research.

[0005] According to a first aspect of the present disclosure, there is provided a method for generating a research report, including:

[0006] In response to a keyword input by a user, obtaining data information related to the keyword from a preset data source;

[0007] Based on a preset screening loss function, screening the content associated with the keyword in the data information;

[0008] Generating a research report related to the keyword based on the content associated with the keyword.

[0009] Optionally, the preset loss function is determined based on the similarity between the keyword and the data information, the quality score of the preset data source, and the relevance between the keyword and the entities in the corresponding domain knowledge graph;

[0010] The preset loss function L filter is:

[0011] L filter = α·sim(K, D)+β·score(D)+γ·∑rel(K, c)

[0012] Among them, sim(K, D) is the similarity between keyword K and data information D, and score(D i ) is the quality score of the preset data source D i . ∑rel(K, c) is the relevance between keyword K and entity c in the corresponding domain knowledge graph, and α, β, and γ are weight parameters.

[0013] Optionally, the similarity sim(K, D) between the keyword and the data information is:

[0014]

[0015] Among them, A i is the vector corresponding to the text of the keyword, and B i is the vector corresponding to the text of the data information.

[0016] Optionally, the quality score score(D i ) of the preset data source D i is:

[0017] score(D i ) = ω 1 ·trust(D i ) + ω 2 ·freq(D i ) + ω 3 ·comp(D i )

[0018] Among them, trust(D i ) is the data source credibility of the preset data source D i , freq(D i ) is the update frequency of the preset data source D i , and comp(D i ) represents the data integrity of the preset data source D i .

[0019] Optionally, the relevance rel(K, c) between the keyword and any entity in the corresponding domain knowledge graph is:

[0020]

[0021] Among them, d(K, c) is the shortest path length between the node of keyword K in the knowledge graph and the node of entity c in the knowledge graph.

[0022] Optionally, the method further includes: in response to the keyword input by the user, obtaining the knowledge graph information of the domain corresponding to the keyword, where the knowledge graph information includes entity information and relationship information.

[0023] Optionally, a research report related to the keyword is generated based on the content associated with the keyword, including:

[0024] In response to the user's input, determine report generation parameters, which are used to determine the depth and breadth of the content of the research report;

[0025] Generate a research report related to the keyword according to the report generation parameters and the content associated with the keyword.

[0026] According to a second aspect of the present disclosure, there is provided a device for generating a research report, including:

[0027] A response module, configured to obtain data information related to the keyword from a preset data source in response to the keyword input by the user;

[0028] A screening module, configured to screen the content associated with the keyword in the data information based on a preset screening loss function;

[0029] A generation module, configured to generate a research report related to the keyword based on the content associated with the keyword.

[0030] According to a third aspect of the present disclosure, there is provided an electronic device, including a processor and a memory, where computer instructions are stored in the memory, and when the computer instructions are executed by the processor, the steps of the method described in the first aspect are implemented.

[0031] According to a fourth aspect of the present disclosure, there is provided a storage medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0032] In this example, a method for generating a research report is provided. Data information can be automatically obtained from a preset information source through the keyword input by the user, and the content associated with the keyword is further screened based on a preset loss function, improving the accuracy of the research report, thereby generating a corresponding research report. Compared with the manual research method, the efficiency is greatly enhanced, a large amount of repetitive labor is avoided, and the problem of easily leading to errors is avoided.

[0033] Through the following detailed description of the exemplary embodiments of the present disclosure with reference to the accompanying drawings, other features and advantages of the embodiments of the present disclosure will become clear. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The drawings forming a part of the specification depict the embodiments of the present disclosure and, together with the description, are used to explain the principles of the embodiments of the present disclosure.

[0035] Figure 1 is a flowchart of a method for generating a research report according to an embodiment;

[0036] Figure 2Schematic diagram of a research report generation device according to an embodiment;

[0037] Figure 3 Schematic diagram of the structure of an electronic device according to an embodiment. Detailed implementation manners

[0038] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that: Unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions, and values set forth in these embodiments do not limit the scope of the present invention.

[0039] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way a limitation on the present invention, its application, or use.

[0040] Techniques and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, such techniques and devices should be considered as part of the specification.

[0041] In all the examples shown and discussed herein, any specific values should be construed as merely exemplary and not as limitations. Therefore, other examples of the exemplary embodiments may have different values.

[0042] It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0043] It should be noted that all actions related to the collection, storage, use, processing, transmission, provision, disclosure, deletion, etc. of data in the present disclosure are carried out on the premise of complying with relevant data protection regulations and policies of the country or region where it is located and with the full authorization of the corresponding data owners.

[0044] The embodiment of the present application discloses a method for generating a research report, as Figure 1 shown, including steps S11 - S13.

[0045] Step S11, in response to a keyword input by a user, obtain data information related to the keyword from a preset data source.

[0046] In this embodiment, when generating a research report, the user can input a keyword for the research report. In this example, it can be in the form of a single word, for example, new energy, biology, etc., or a keyword composed of multiple words, such as "new energy battery performance parameters", etc.

[0047] In this embodiment, the preset data source can be set according to the field. For example, for the academic field, the preset information sources can be various academic websites, forums, or platforms, etc. For the commercial field, the corresponding field web pages, etc. can be set as the preset information sources. In this embodiment, the HTML structure can be identified and parsed through web crawler technology and corresponding parsing tools, and the system can effectively locate the links and files storing text information, so as to obtain the data information in the preset information sources.

[0048] In this embodiment, the data information includes text content and other data resource content. In this example, documents or reports containing keywords can be retrieved in the preset information sources, and the text content or data content therein can be extracted as the data information related to the keywords. In another example, the relevance between the keywords and the title or abstract of a certain document or report can be calculated to determine the data content associated with the keywords. In this example, the method for determining the relevance between the keywords and the document report can be determined by calculating the cosine similarity of text vectors.

[0049] Step S12, screen the content associated with the keywords in the data information based on the preset screening loss function.

[0050] In an example, the data information can be input into the RAG (Retrieval-Augmented Generation) module of the natural language model as the basic data set, and the RAG module can screen the collected data information to ensure that the content generated in the report closely revolves around the keywords. Then, a screening loss function can be set according to the requirements to perform pruning data processing on all relevant content to obtain relevant content.

[0051] In an example of this embodiment, the preset loss function is determined based on the similarity between the keywords and the data information, the quality score of the preset data source, and the relevance between the keywords and the entities in the knowledge graph of the corresponding field.

[0052] For the accuracy of the final generated research report, when setting the preset loss function, multi-dimensional considerations can be made. First, it is necessary to combine the similarity between the keywords and the data information, that is, their association relationship, to avoid excessive content unrelated to the keywords in the data information. At the same time, the quality score of the preset data source can also be considered. For example, for the data information, the authenticity and timeliness of different data sources will vary, and users can preset the quality score of each preset information source to enhance the reliability and timeliness of the generated research report. In addition, the relevance between the keywords and the entities in the knowledge graph of the corresponding field can also be considered.

[0053] In an example of this embodiment, in response to a keyword input by a user, knowledge graph information in the field corresponding to the keyword is obtained, and the knowledge graph information includes entity information and relationship information.

[0054] After the user inputs a keyword, the field corresponding to the keyword can be determined first, and then the knowledge graph information in the corresponding field can be obtained. In one example, a knowledge graph can be constructed through the text information in the knowledge base of the corresponding field. First, knowledge extraction can be performed on the text information in the knowledge base to extract entities, relationships, attributes, etc. in the text information. Taking the knowledge base in the field of papers as an example, the extracted entities can include authors, papers, institutions, years, journals, etc., and then the relationships between the entities are determined, such as author - published paper, etc. The extraction of entities, relationships, and attributes in the text information can be identified through a deep learning model, such as the BERT - BiLSTM - CRF model, or can be extracted through regular expressions or template matching. Then, the extracted entities and relationships are fused, including alignment and deduplication, to ensure that the same entity appears only once in the knowledge graph. The unification of entities is achieved through string similarity and semantic similarity calculations. The problems of repetition and ambiguity in the extraction results are solved. The fused knowledge is stored in a structured manner to support efficient querying and operations. The processed entities and relationships are stored in the knowledge base to formally form a knowledge graph. Then, the corresponding knowledge graph information, including information on entities and relationships, etc., is extracted. Then, the relevance between the keyword and each entity can be calculated, and the relevance between the keyword and the corresponding field can be comprehensively evaluated to enhance the reliability of the generated research report.

[0055] In another example, the relevance between the keyword and the entities in the knowledge graph of the corresponding field can be the relevance between the keyword and the entities included in the data information in the knowledge graph of the corresponding field. The entities and relationships included in the data information can also be determined through the method of knowledge extraction.

[0056] Step S13, generate a research report related to the keyword based on the content associated with the keyword.

[0057] In this example, a large - language model can be used to generate a corresponding research report based on the content associated with the keyword. Then, the model can also be used to perform consistency checking and content refreshing on the generated research report to improve the timeliness and consistency of the content.

[0058] In this example, a method for generating a research report is provided. Data information can be automatically obtained from a preset information source through a keyword input by a user, and the content associated with the keyword can be further screened based on a preset loss function to improve the accuracy of the research report, thereby generating a corresponding research report. Compared with the manual research method, the efficiency is greatly enhanced, and a large amount of repetitive labor is avoided, which is extremely prone to errors.

[0059] In an example of this embodiment, the preset loss function L filter is:

[0060] L filter = α·sim(K, D) + β·score(D) + γ·∑rel(K, c)

[0061] where sim(K, D) is the similarity between the keyword K and the data information D, score(D i ) is the quality score of the preset data source D i , and ∑rel(K, c) is the relevance between the keyword K and the entity c in the corresponding domain knowledge graph. α, β, and γ are weight parameters.

[0062] In this example, the preset loss function can be determined based on the similarity between the keyword and the data information, the quality score of the preset data source, the relevance between the keyword and the entity in the corresponding domain knowledge graph, and the corresponding weight parameters. The similarity between the keyword and the data information can be determined by calculating the cosine similarity of the text vectors. In this example, the text in the keyword and the corresponding data information can be first converted into vector forms, and then the cosine similarity between the two can be calculated.

[0063] In an example of this embodiment, the similarity sim(K, D) between the keyword and the data information is:

[0064]

[0065] where A i is the vector corresponding to the text of the keyword, and B i is the vector corresponding to the text of the data information.

[0066] In an example of this embodiment, the quality score score(D i ) of the preset data source D i is:

[0067] score(D i ) = ω 1 ·trust(D i ) + ω 2 ·freq(D i ) + ω 3 ·comp(D i )

[0068] where trust(D i ) is the data source credibility of the preset data source D i , freq(D i ) is the frequency of the preset data source D iThe update frequency, comp(D i ) represents the data integrity of the preset data source D i .

[0069] In this example, the credibility, update frequency, and data integrity of the data source can be pre-acquired by the user. In another example, after the research report is generated, the credibility and integrity of the data source can also be automatically updated in response to the user's evaluation input for the research report. For the update frequency of the data source, it can be automatically obtained by monitoring the link of the data source.

[0070] In an example of this embodiment, the relevance rel(K, c) between a keyword and any entity in the corresponding domain knowledge graph is as follows:

[0071]

[0072] where d(K, c) is the shortest path length between the node of the keyword K in the knowledge graph and the node of the entity c in the knowledge graph. After determining the relevance between the keyword and each entity, their sum value can be used as the relevance between the keyword and the entity c in the corresponding domain knowledge graph. In another example, the relevance between the keyword and the entity in the corresponding domain knowledge graph included in the acquired data information can be determined.

[0073] In an example of this embodiment, generating a research report related to the keyword based on the content associated with the keyword includes: in response to the user's input, determining report generation parameters, where the report generation parameters are used to determine the depth and breadth of the content of the research report; generating a research report according to the report generation parameters and the content associated with the keyword.

[0074] In an example, when generating a research report, the user can input the requirements of their own research report or directly set the report generation parameters. In this example, the report generation parameters can be used to adjust the depth and breadth of the content of the research report. For example, when the user is more inclined to generate a research report with a larger breadth, the natural language processing model can extract the abstract of each content associated with the keyword respectively, and process all these abstracts to generate a research report. In another example, when the user is more inclined to generate a research report with a greater depth, the natural language processing model can process one or several contents most associated with the keyword to obtain a more specific and detailed research report. Even further, the second keyword in the content can be extracted, and steps S11 and S12 can be repeatedly executed through the second keyword to obtain content with a greater depth, so as to generate a research report.

[0075] In another example, a human-computer interaction interface for the user to input research report parameters or select a research report mode can be provided, enabling the user to directly adjust the research report generation parameters or adjust them by selecting a mode. For example, a depth report mode or a breadth report mode can be provided in the interface for the user to choose. After the user makes a selection, the research report parameters can be automatically adjusted.

[0076] In another embodiment, a generation device 100 for a research report is further provided, as Figure 2 shown, including: a response module 101, configured to obtain data information related to a keyword from a preset data source in response to the keyword input by the user; a screening module 102, configured to screen the content associated with the keyword in the data information based on a preset screening loss function; and a generation module 103, configured to generate a research report related to the keyword based on the content associated with the keyword.

[0077] Optionally, the preset loss function is determined based on the similarity between the keyword and the data information, the quality score of the preset data source, and the relevance between the keyword and the entities in the corresponding domain knowledge graph.

[0078] Optionally, the preset loss function L filter is:

[0079] L filter = α·sim(K, D)+β·score(D)+γ·∑rel(K, c)

[0080] where sim(K, D) is the similarity between the keyword K and the data information D, score(D i ) is the quality score of the preset data source D i , ∑rel(K, c) is the relevance between the keyword K and the entity c in the corresponding domain knowledge graph, and α, β, and γ are weight parameters.

[0081] Optionally, the similarity sim(K, D) between the keyword and the data information is:

[0082]

[0083] where A i is the vector corresponding to the text of the keyword, and B i is the vector corresponding to the text of the data information.

[0084] Optionally, the quality score score(D i ) of the preset data source D i is:

[0085] score(D i ) = ω 1 ·trust(D i) + ω 2 · freq(D i ) + ω 3 · comp(D i )

[0086] where trust(D i ) is the data source credibility of the preset data source D i freq(D i ) is the update frequency of the preset data source D i comp(D i ) represents the data integrity of the preset data source D i .

[0087] Optionally, the relevance rel(K, c) between the keyword and any entity in the corresponding domain knowledge graph is:

[0088]

[0089] where d(K, c) is the length of the shortest path between the node of the keyword K in the knowledge graph and the node of the entity c in the knowledge graph.

[0090] Optionally, the device further includes: obtaining knowledge graph information corresponding to the keyword in response to the keyword input by the user, where the knowledge graph information includes entity information and relationship information.

[0091] Optionally, the generating module is specifically configured to: determine report generation parameters in response to the user input, where the report generation parameters are used to determine the depth and breadth of the content of the research report; generate a research report according to the report generation parameters and the content associated with the keyword.

[0092] As Figure 3 shown, an embodiment of the present application further provides an electronic device 200, including a processor 201 and a memory 202, where computer instructions are stored in the memory 202, and when the computer instructions are executed by the processor 201, the steps of any one of the methods in the embodiment of the research report generation method are implemented.

[0093] An embodiment of the present application further provides a storage medium, on which computer instructions are stored, and when the computer instructions are executed by the processor, any one of the above-mentioned embodiments of the research report generation is implemented, and the same technical effects can be achieved. To avoid repetition, it will not be elaborated here.

[0094] The various embodiments in the present disclosure are described in a progressive manner. For the same or similar parts among the various embodiments, reference can be made to each other, and the key points of each embodiment are the differences from other embodiments. In particular, for device and equipment embodiments, since they are basically similar to method embodiments, the description is relatively simple, and for the relevant parts, reference can be made to the description in the method embodiments.

[0095] The above describes specific embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0096] Embodiments of the present disclosure may be systems, methods, and / or computer program products. A computer program product may include a computer-readable storage medium having thereon computer-readable program instructions for causing a processor to implement various aspects of the embodiments of the present disclosure.

[0097] A computer-readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punched card or raised structures in a groove having instructions stored thereon, and any suitable combination of the foregoing. The computer-readable storage medium used herein is not to be construed as a transitory signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0098] The computer-readable program instructions described herein can be downloaded to various computing / processing devices from a computer-readable storage medium or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0099] The computer program instructions for performing the operations of the embodiments of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state-setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages - such as Smalltalk, C++, etc., and conventional procedural programming languages - such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or, alternatively, may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the embodiments of the present disclosure.

[0100] Aspects of the embodiments of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer-readable program instructions.

[0101] These computer-readable program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus to produce a machine such that the instructions, when executed by the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions may also be stored in a computer-readable storage medium that causes a computer, a programmable data processing apparatus, and / or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises a manufacture including instructions for implementing various aspects of the functions / acts specified in one or more blocks of the flowchart and / or block diagram.

[0102] The computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process such that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / acts specified in one or more blocks of the flowchart and / or block diagram.

[0103] The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of code, or a portion of an instruction, and the module, segment of code, or portion of an instruction contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block of the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or acts, or by a combination of dedicated hardware and computer instructions. It is well known to those skilled in the art that implementation by hardware, implementation by software, and implementation by a combination of software and hardware are equivalent.

[0104] The embodiments of the present disclosure have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or the improvement of technologies in the market, or to enable other ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for generating a research report, characterized in that: include: In response to a keyword input by a user, acquiring data information related to the keyword from a preset data source; Filtering the content associated with the keyword in the data information based on a preset filtering loss function; Generate a research report related to the keyword based on the content associated with the keyword.

2. The method according to claim 1, characterized in that: The preset loss function is determined based on the similarity between the keyword and the data information, the quality score of the preset data source, and the relevance between the keyword and the entity in the corresponding domain knowledge graph; The preset loss function L filter for: L filter =α·sim(K,D)+β·score(D)+γ·∑rel(K,c) Among them, sim(K,D) is the similarity between the keyword K and the data information D, score(D i ) The preset data source D i is the quality score of the keyword K, ∑rel(K,c) is the correlation between the keyword K and the entity c in the corresponding domain knowledge graph, and α, β, and γ are weight parameters.

3. The method according to claim 2, characterized in that The similarity sim(K, D) between the keyword and the data information is: Among them, A i is the vector corresponding to the text of the keyword, B i A vector corresponding to the text of the data information.

4. The method according to claim 3, characterized in that The preset data source D i The quality score (D i )for: score(D i )=ω1·trust(D i )+ω2·req(D i )+ω3·comp(D i ) Among them, trust(D i ) is the preset data source D i The credibility of the data source, freq(D i ) is the preset data source D i The update frequency of comp(D i ) represents the preset data source D i data integrity.

5. The method according to claim 4, characterized in that The correlation rel(K,c) between the keyword and any entity in the corresponding domain knowledge graph is: Among them, d(K,c) is the shortest path length between the node of keyword K in the knowledge graph and the node of entity c in the knowledge graph.

6. The method according to claim 5, characterized in that The method further comprises: In response to keywords input by the user, knowledge graph information of the field corresponding to the keyword is obtained, and the knowledge graph information includes entity information and relationship information.

7. The method according to claim 1, characterized in that The generating of a research report related to the keyword based on the content associated with the keyword includes: In response to user input, determining report generation parameters, wherein the report generation parameters are used to determine the depth and breadth of the content of the research report; The research report is generated according to the report generation parameters and the content associated with the keywords.

8. A device for generating a research report, characterized in that: include: A response module, used to obtain data information related to the keyword from a preset data source in response to the keyword input by the user; A screening module, used for screening the content associated with the keyword in the data information based on a preset screening loss function; A generation module is used to generate a research report related to the keyword based on the content associated with the keyword.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer instructions, and when the computer instructions are executed by the processor, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images