A method, device and medium for expanding ideas based on natural language
By constructing keyword feature vectors and using filtering algorithms to recommend similar keywords, the difficulties researchers face in discovering new research topics have been resolved, enabling rapid expansion of ideas and the discovery of new topics.
Patent Information
- Application Number
- CN202111645865.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-29
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-12-29
AI Technical Summary
Discovering new research topics in existing technologies is quite difficult. Researchers who focus on a specific research direction for a long time are prone to developing fixed mindsets and cannot easily extend their thinking to other new topics related to their own research, resulting in a long time cycle for discovering new topics.
By obtaining a text set of a specified user, using a comprehensive thesaurus and word segmentation algorithm to perform word segmentation, constructing keyword feature vectors, and using a preset filtering algorithm to calculate the relevance coefficient distance with other users, keywords similar to the specified user's research direction are recommended to broaden research ideas.
It effectively shortens the time to discover new research topics, helps researchers expand their research ideas through recommended keywords, avoids fixed mindsets, and improves the efficiency of new topic discovery.
Smart Images

Figure CN114298036B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data mining technology, and in particular to a method, device and medium for expanding ideas based on natural language. Background Art
[0002] At present, the exploration of new research topics is of great significance to both academia and industry. For academia, conducting research on new topics helps to continue the scientific research process or open up scientific research directions, and promote the generation of new scientific research results; for industry, research on new topics can promote the development and production of new products, solve technical problems for enterprises or bring new profit points. At the same time, it can also demonstrate the core scientific and technological strength of enterprises through the publication of papers and patents, protect the rights and interests of enterprises, and enhance the competitiveness of enterprises.
[0003] However, for the author, the discovery of new research topics is relatively scarce. Usually, the discovery is outside the regular research topics and the idea that suddenly flashes at a certain moment, or the new ideas of continuity or cooperation generated by long-term research in a certain direction and seminar exchanges with researchers in similar industries. Summary of the invention
[0004] The embodiments of the present application provide a method, device and medium for expanding ideas based on natural language, which are used to solve the problem that it is difficult to discover new research topics in the prior art. The time period for discovery by researchers themselves is long. Researchers who conduct research in a specific research direction for a long time are prone to form a fixed mindset and cannot easily expand their ideas to other new topics related to their own research topics.
[0005] On the one hand, an embodiment of the present application provides a method for expanding ideas based on natural language, including: obtaining a text set corresponding to a specified user, and based on a comprehensive vocabulary, segmenting the text in the text set through a preset segmentation algorithm to obtain corresponding segmentation results; wherein the text set includes all documents published by the specified user; processing the segmentation results through a preset statistical algorithm to obtain keywords in the text set; determining corresponding weights according to the specified keywords, and constructing a keyword feature vector corresponding to the specified user according to the corresponding weights; based on a pre-acquired keyword feature matrix and a preset filtering algorithm, determining keywords recommended for the specified user according to the keyword feature vector corresponding to the specified user, and recommending the keywords to the specified user; wherein the keyword feature matrix is composed of the splicing of keyword feature vectors corresponding to several users.
[0006] In one implementation of the present application, before the text in the text collection is segmented by a preset segmentation algorithm to obtain the corresponding segmentation results, it also includes: obtaining papers corresponding to several users from the Internet, and extracting keywords in several papers to build a keyword vocabulary based on the keywords in the several papers; and obtaining text collections corresponding to several users from the Internet, and obtaining academic terms from several text collections through a new word discovery algorithm to build a new word vocabulary based on the academic terms; and obtaining a basic vocabulary from the Internet, and combining the basic vocabulary, the keyword vocabulary and the new word vocabulary into a comprehensive vocabulary.
[0007] In one implementation of the present application, before determining the recommended keywords for the specified user based on the keyword feature vector matrix obtained in advance and the keyword feature vector corresponding to the specified user, the method also includes: obtaining a text set corresponding to the user, and segmenting the text set corresponding to the user through the comprehensive vocabulary to obtain a segmentation result of the text set corresponding to the user; constructing a keyword feature vector corresponding to the user based on the segmentation result; adding the weighted values of the specified keywords in the text set corresponding to the user to obtain the dimension of the user in the keyword feature vector, and concatenating the keyword feature vectors corresponding to several users to obtain a keyword feature matrix.
[0008] In one implementation of the present application, before determining the recommended keywords for the designated user, it also includes: obtaining keyword feature vectors corresponding to several users from a pre-acquired keyword feature matrix; and calculating the correlation coefficient distance between the keyword feature vector corresponding to the designated user and the keyword feature vectors corresponding to the several users through a preset filtering algorithm.
[0009] In one implementation of the present application, the correlation coefficient distance calculation formula between A and B is: Among them, A and B are two different authors, and the keyword feature vector corresponding to author A is X A =(x A1 ,x A2 ,…,x An ), the keyword feature vector corresponding to author B is X B =(x B1 ,x B2 ,…,x Bn ).
[0010] In one implementation of the present application, based on the pre-acquired keyword feature matrix and the preset filtering algorithm, the recommended keywords for the designated user are determined according to the keyword feature vector corresponding to the designated user, specifically including: according to the correlation coefficient distance between the designated user and the plurality of users, obtaining a preset number of users whose correlation coefficient distance with the designated user is less than a preset distance threshold from the plurality of users; according to the correlation coefficient between the designated user and the plurality of users, performing weighted summation of the keyword feature vectors corresponding to the preset number of users to obtain a corresponding keyword weight vector.
[0011] In one implementation of the present application, the keyword weight vector calculation formula is: Among them, m is the preset number corresponding to the existing authors.
[0012] In one implementation of the present application, it also includes: according to the corresponding keyword weight vector, from the keywords corresponding to the several users, determining a preset number of keywords whose corresponding keyword weight vectors are less than a preset weight vector threshold; using the determined keywords as recommended words, and recommending them to the designated user, so that the designated user can expand his thinking based on the recommended words.
[0013] On the other hand, an embodiment of the present application also provides a natural language-based idea expansion device, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain a text set corresponding to a specified user, and based on a comprehensive vocabulary, segment the text in the text set through a preset segmentation algorithm to obtain a corresponding segmentation result; wherein the text set includes all documents published by the specified user; the segmentation result is processed through a preset statistical algorithm to obtain keywords in the text set; corresponding weights are determined according to the specified keywords, and a keyword feature vector corresponding to the specified user is constructed according to the corresponding weights; based on a pre-acquired keyword feature matrix and a preset filtering algorithm, the keywords recommended for the specified user are determined according to the keyword feature vector corresponding to the specified user, and the keywords are recommended to the specified user; wherein the keyword feature matrix is composed of the splicing of keyword feature vectors corresponding to several users.
[0014] On the other hand, an embodiment of the present application also provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to: obtain a text set corresponding to a specified user, and based on a comprehensive vocabulary, segment the text in the text set through a preset segmentation algorithm to obtain a corresponding segmentation result; wherein the text set includes all documents published by the specified user; process the segmentation result through a preset statistical algorithm to obtain keywords in the text set; determine corresponding weights based on the specified keywords, and construct a keyword feature vector corresponding to the specified user based on the corresponding weights; based on a pre-acquired keyword feature matrix and a preset filtering algorithm, determine the keywords recommended for the specified user based on the keyword feature vector corresponding to the specified user, and recommend the keywords to the specified user; wherein the keyword feature matrix is composed of the splicing of keyword feature vectors corresponding to several users.
[0015] The embodiments of the present application provide a method, device and medium for expanding ideas based on natural language, which at least include the following beneficial effects: using a preset word segmentation algorithm based on a comprehensive vocabulary to segment text in a text set corresponding to a specified user, and then processing the word segmentation results through a preset statistical method, so that all keywords corresponding to the specified user can be determined; corresponding weights can be determined according to the specified keywords, so that a keyword feature vector corresponding to the specified user can be constructed according to the corresponding weights; using a preset filtering algorithm, the distance between the keyword feature vector corresponding to the specified user and the keyword feature vectors corresponding to other users in a pre-acquired keyword feature matrix is calculated, so that keywords that are relatively similar to the research direction of the specified user can be determined through the distance between other users and the specified user; by recommending the determined keywords to the specified user, the specified user can expand ideas according to the recommended keywords and determine new research topics. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0017] Figure 1 A flowchart of a natural language-based method for expanding ideas provided in an embodiment of the present application;
[0018] Figure 2 A schematic diagram of keyword extraction provided in an embodiment of the present application;
[0019] Figure 3 A schematic diagram of constructing a keyword feature vector provided in an embodiment of the present application;
[0020] Figure 4 A schematic diagram of the internal structure of a natural language-based idea expansion device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0021] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0022] The embodiment of the present application provides a method, device and medium for expanding ideas based on natural language, which uses a preset word segmentation algorithm based on a comprehensive word library to segment the text in the text set corresponding to the specified user, and then processes the word segmentation results through a preset statistical method, so that all keywords corresponding to the specified user can be determined; the corresponding weight can be determined according to the specified keyword, so that the keyword feature vector corresponding to the specified user can be constructed according to the corresponding weight; the keyword feature vector corresponding to the specified user is calculated through a preset filtering algorithm, and the distance between the keyword feature vector corresponding to other users in the pre-acquired keyword feature matrix is calculated, so that the keyword that is similar to the research direction of the specified user can be determined through the distance between other users and the specified user; by recommending the determined keywords to the specified user, the specified user can expand ideas according to the recommended keywords. The technical problem that it is difficult to discover new research topics in the prior art, the time period for discovery by scientific researchers is long, and it is easy to form a fixed mindset for scientific researchers to conduct research on a specific research direction for a long time, and it is difficult to easily expand ideas to other new topics related to their own research topics is solved.
[0023] The technical solution proposed in the embodiments of the present application is described in detail below with reference to the accompanying drawings.
[0024] Figure 1 A flowchart of a method for expanding ideas based on natural language is provided in an embodiment of the present application. Figure 1 As shown, the natural language-based idea expansion method provided in the embodiment of the present application mainly includes the following steps:
[0025] Step 101: Acquire a text set corresponding to a specified user, and segment the text in the text set based on a comprehensive word library using a preset segmentation algorithm to obtain a corresponding segmentation result.
[0026] The server obtains all documents, papers, articles and other contents published by the specified user from the network, and combines them to form a text set corresponding to the specified user. Then, based on the comprehensive vocabulary and through the preset word segmentation algorithm, the text in the text set corresponding to the specified user is segmented and the corresponding word segmentation results are obtained.
[0027] It should be noted that when obtaining the text set corresponding to a specified user, the embodiment of the present application needs to construct the text set as extensively as possible, so that the keywords associated with the specified user can be determined based on a sufficient number of documents, papers, articles, etc., so that the feature vector corresponding to a single specified user can be as complete as possible.
[0028] In one embodiment of the present application, the server uses a preset word segmentation algorithm to segment text in a text set. Before obtaining the corresponding word segmentation results, it is necessary to obtain papers corresponding to several users from the network and extract keywords from the papers, thereby building a keyword thesaurus based on the keywords in the extracted papers; and obtain text sets corresponding to several users from the network, and obtain academic terms from the text sets through a new word discovery algorithm, thereby building a new word thesaurus based on the academic terms; and obtain a basic thesaurus from the network, and combine the basic thesaurus, the keyword thesaurus and the new word thesaurus into a comprehensive thesaurus, so that the server can segment the text set corresponding to the specified user according to the words in the comprehensive thesaurus and extract the keywords in the text set. This can enhance the accuracy and professionalism of the word segmentation results.
[0029] Figure 2 This is a schematic diagram of keyword extraction provided in the embodiment of the present application. Figure 2 As shown, the comprehensive word library of the embodiment of the present application is composed of a basic word library, a keyword word library and a new word word library. Based on the comprehensive word library, the present application performs word segmentation on the text set corresponding to the user through a preset word segmentation algorithm, thereby obtaining the keywords corresponding to the user.
[0030] Step 102: Process the word segmentation results using a preset statistical algorithm to obtain keywords in the text collection.
[0031] After processing the word segmentation results obtained by using a preset statistical algorithm, the server can obtain the keywords in the text set corresponding to the specified user, so that the research direction corresponding to the specified user can be determined based on the keywords corresponding to the specified user.
[0032] Step 103: Determine a corresponding weight according to the specified keyword, and construct a keyword feature vector corresponding to the specified user according to the corresponding weight.
[0033] The server determines the weight of the specified keyword in the specified user according to the specified keyword, and thus constructs a corresponding keyword feature vector for the specified user according to the weight of each keyword of the specified user.
[0034] In one embodiment of the present application, before the server determines the recommended keywords for the specified user based on the keyword feature vector matrix obtained in advance and the preset filtering algorithm, according to the keyword feature vector corresponding to the specified user, it is also necessary to obtain the text sets corresponding to other users, and then segment the text set corresponding to the user based on the comprehensive vocabulary through the preset word segmentation algorithm, so as to obtain the word segmentation result of the text set corresponding to the user; construct the keyword feature vector corresponding to the user according to the weight corresponding to the keyword of the user in the word segmentation result; then add the weighted values of the specified keywords in the text set corresponding to the user to obtain the dimension of the keyword feature vector of the user. It should be noted that the dimension of the keyword feature vector corresponding to each user is the same, and the dimension is the number of keywords. Then the keyword feature vectors corresponding to several users are spliced to obtain the keyword feature matrix corresponding to several users.
[0035] Figure 3 This is a schematic diagram of the construction of a keyword feature vector provided in the embodiment of the present application. Figure 3 As shown, the embodiment of the present application obtains the keyword corresponding to the user based on the word segmentation result of the text set corresponding to the user, and then adds the TF-IDF value of the keyword in the text set corresponding to the user to determine the dimension of the user in the keyword feature vector, thereby constructing a key feature vector for the user.
[0036] Step 104: Based on the pre-acquired keyword feature matrix and the preset filtering algorithm, according to the keyword feature vector corresponding to the designated user, determine the keywords recommended for the designated user, and recommend the keywords to the designated user.
[0037] The server determines the keyword feature vectors corresponding to several users from the pre-acquired keyword feature matrix, and calculates the distance between the keyword feature vector corresponding to the specified user and the keyword feature vectors corresponding to several users in the keyword feature matrix through a preset filtering algorithm, so as to find keywords similar to the research direction of the specified user, and use them as keywords recommended for the specified user. Then, the keyword is recommended to the specified user, providing the specified user with a source of inspiration for expanding his or her ideas, and enabling the specified user to expand his or her research ideas through the keyword and determine a new research topic.
[0038] In one embodiment of the present application, before determining the recommended keywords for a specified user, the server obtains keyword feature vectors corresponding to several users from a pre-acquired keyword feature matrix, and then calculates the correlation coefficient distance between the keyword feature vector corresponding to the specified user and the keyword feature vectors corresponding to several users through a preset filtering algorithm.
[0039] It should be noted that the preset filtering algorithm in the embodiment of the present application is a collaborative filtering algorithm. The present application uses the collaborative filtering algorithm to recommend specific keywords to designated users as a source of inspiration for expanding ideas. Keyword recommendation is applicable to any collaborative filtering algorithm. For the sake of simple description, the present application only uses the user-based collaborative filtering algorithm as an example for explanation.
[0040] Specifically, when calculating the distance between the keyword feature vector corresponding to a specified user and the keyword feature vectors corresponding to other users, the calculation formula of the correlation coefficient distance is: The present application can calculate the distance between the designated user A and other user B by inputting the keyword feature vector corresponding to the designated user A and the keyword feature vector corresponding to other user B into the above calculation formula.
[0041] It should be noted that A in the embodiment of the present application represents a designated user, B represents a user in the keyword feature matrix, and the keyword feature vector corresponding to the designated user A is X A =(x A1 ,x A2 ,…,x An ), the keyword feature vector corresponding to other user B is X B =(x B1 ,x B2 ,…,x Bn ).
[0042] In one embodiment of the present application, the server obtains a preset number of users whose distance from the designated user is less than a preset distance threshold from the designated user from the plurality of users based on the distance between the designated user and the plurality of users; then, based on the correlation coefficient between the designated user and the plurality of users, the server performs weighted summation on the keyword feature vectors corresponding to the preset number of users to obtain a corresponding keyword weight vector.
[0043] Specifically, in the embodiment of the present application, when calculating the keyword weight vectors corresponding to a preset number of users, the formula for calculating the keyword weight vector is:
[0044] It should be noted that m in the embodiment of the present application represents the preset number of users for which weighted summation of keyword feature vectors is required.
[0045] In one embodiment of the present application, the server determines keywords whose keyword weight vectors of a preset number of keywords are less than a preset weight vector threshold from the keywords corresponding to a number of users according to the keyword weight vectors corresponding to a preset number of users; and then recommends the determined keywords as recommended words to the designated user, so that the designated user can expand his ideas according to the recommended words, thereby determining a new research topic. The present application can solve the problem by converting the abstract idea expansion into concrete keyword recommendation, and then using technical means.
[0046] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, the embodiment of this application also provides a natural language-based thinking expansion device, whose structure is as follows Figure 4 shown.
[0047] Figure 4 The internal structure diagram of a natural language-based idea expansion device provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the device includes at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain a text set corresponding to a specified user, and based on a comprehensive vocabulary, segment the text in the text set through a preset segmentation algorithm to obtain a corresponding segmentation result; wherein the text set includes all documents published by the specified user; the segmentation result is processed through a preset statistical algorithm to obtain keywords in the text set; the corresponding weight is determined according to the specified keyword, and the keyword feature vector corresponding to the specified user is constructed according to the corresponding weight; based on the pre-acquired keyword feature matrix and the preset filtering algorithm, the keyword recommended for the specified user is determined according to the keyword feature vector corresponding to the specified user, and the keyword is recommended to the specified user; wherein the keyword feature matrix is composed of the splicing of the keyword feature vectors corresponding to several users.
[0048] The embodiment of the present application also provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to: obtain a text set corresponding to a specified user, and based on a comprehensive vocabulary, segment the text in the text set through a preset segmentation algorithm to obtain a corresponding segmentation result; wherein the text set includes all documents published by the specified user; process the segmentation result through a preset statistical algorithm to obtain keywords in the text set; determine corresponding weights according to the specified keywords, and construct a keyword feature vector corresponding to the specified user according to the corresponding weights; based on a pre-acquired keyword feature matrix and a preset filtering algorithm, determine the keywords recommended for the specified user according to the keyword feature vector corresponding to the specified user, and recommend the keywords to the specified user; wherein the keyword feature matrix is composed of the splicing of keyword feature vectors corresponding to several users.
[0049] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.
[0050] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.
[0051] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0052] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0053] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0054] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0055] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0056] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0057] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0058] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0059] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims
1. A method for expanding ideas based on natural language, characterized in that: The method comprises: Acquire a text set corresponding to a specified user, and segment the text in the text set based on a comprehensive word library and a preset word segmentation algorithm to obtain a corresponding word segmentation result; wherein the text set includes all documents published by the specified user; Processing the word segmentation results by a preset statistical algorithm to obtain keywords in the text set; Determine a corresponding weight according to the specified keyword, and construct a keyword feature vector corresponding to the specified user according to the corresponding weight; Based on the pre-acquired keyword feature matrix and the preset filtering algorithm, determine the keywords recommended for the designated user according to the keyword feature vector corresponding to the designated user, and recommend the keywords to the designated user; The keyword feature matrix is composed of the keyword feature vectors corresponding to several users; Before determining the keywords recommended for the designated user, the method further includes: Obtain keyword feature vectors corresponding to a number of users from a pre-acquired keyword feature matrix; By using a preset filtering algorithm, respectively calculating the correlation coefficient distance between the keyword feature vector corresponding to the designated user and the keyword feature vectors corresponding to the plurality of users; The method further comprises: The correlation coefficient distance calculation formula between A and B is: Among them, A represents the specified user, B represents other users, and the keyword feature vector corresponding to the specified user A is X A =(x A1 ,x A2 ,…,x An ), the keyword feature vector corresponding to other user B is X B =(x B1 ,x B2 ,…,x Bn ); The step of determining the recommended keywords for the designated user based on the pre-acquired keyword feature matrix and the preset filtering algorithm according to the keyword feature vector corresponding to the designated user specifically includes: According to the correlation coefficient distance between the designated user and the plurality of users, a preset number of users whose correlation coefficient distance with the designated user is less than a preset distance threshold are obtained from the plurality of users; According to the correlation coefficient between the designated user and the plurality of users, weighted summing is performed on the keyword feature vectors corresponding to the preset number of users to obtain a corresponding keyword weight vector; The method further comprises: The formula for calculating keyword weight vector is: Among them, m is the preset number corresponding to the existing authors.
2. The method for expanding ideas based on natural language according to claim 1, characterized in that: Before segmenting the texts in the text set by a preset segmentation algorithm to obtain corresponding segmentation results, the method further includes: Acquire papers corresponding to a number of users from the Internet, and extract keywords from the papers to build a keyword thesaurus based on the keywords in the papers; and obtaining text collections corresponding to several users from the network, and obtaining academic terms from the several text collections through a new word discovery algorithm, so as to construct a new word database based on the academic terms; And obtain a basic vocabulary from the Internet, and combine the basic vocabulary, the keyword vocabulary and the new word vocabulary into a comprehensive vocabulary.
3. The method for expanding ideas based on natural language according to claim 1, characterized in that: Before determining the keywords recommended for the designated user based on the keyword feature vector matrix obtained in advance and the preset filtering algorithm according to the keyword feature vector corresponding to the designated user, the method further includes: Acquire a text set corresponding to the user, and segment the text set corresponding to the user based on the comprehensive word library by the preset word segmentation algorithm to obtain a word segmentation result of the text set corresponding to the user; Constructing a keyword feature vector corresponding to the user according to the weight corresponding to the user's keyword in the word segmentation result; The weighted values of the specified keywords in the text set corresponding to the user are added to obtain the dimension of the user in the keyword feature vector, and the keyword feature vectors corresponding to several users are spliced to obtain a keyword feature matrix.
4. The method for expanding ideas based on natural language according to claim 1, characterized in that: The method further comprises: According to the corresponding keyword weight vectors, determining, from the keywords corresponding to the plurality of users, a preset number of keywords whose corresponding keyword weight vectors are less than a preset weight vector threshold; The determined keywords are used as recommendation words and recommended to the designated user, so that the designated user can expand his / her thinking according to the recommendation words.
5. A natural language-based thinking expansion device, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire a text set corresponding to a specified user, and segment the text in the text set based on a comprehensive word library and a preset word segmentation algorithm to obtain a corresponding word segmentation result; wherein the text set includes all documents published by the specified user; Processing the word segmentation results by a preset statistical algorithm to obtain keywords in the text set; Determine a corresponding weight according to the specified keyword, and construct a keyword feature vector corresponding to the specified user according to the corresponding weight; Based on the pre-acquired keyword feature matrix and the preset filtering algorithm, determine the keywords recommended for the designated user according to the keyword feature vector corresponding to the designated user, and recommend the keywords to the designated user; The keyword feature matrix is composed of the keyword feature vectors corresponding to several users; Before determining the keywords recommended for the designated user, the method further includes: Obtain keyword feature vectors corresponding to a number of users from a pre-acquired keyword feature matrix; By using a preset filtering algorithm, respectively calculating the correlation coefficient distance between the keyword feature vector corresponding to the designated user and the keyword feature vectors corresponding to the plurality of users; Also includes: The correlation coefficient distance calculation formula between A and B is: Among them, A represents the specified user, B represents other users, and the keyword feature vector corresponding to the specified user A is X A =(x A1 ,x A2 ,…,x An ), the keyword feature vector corresponding to other user B is X B =(x B1 ,x B2 ,…,x Bn ); The step of determining the recommended keywords for the designated user based on the pre-acquired keyword feature matrix and the preset filtering algorithm according to the keyword feature vector corresponding to the designated user specifically includes: According to the correlation coefficient distance between the designated user and the plurality of users, a preset number of users whose correlation coefficient distance with the designated user is less than a preset distance threshold are obtained from the plurality of users; According to the correlation coefficient between the designated user and the plurality of users, weighted summing is performed on the keyword feature vectors corresponding to the preset number of users to obtain a corresponding keyword weight vector; Also includes: The formula for calculating keyword weight vector is: Among them, m is the preset number corresponding to the existing authors.
6. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions are configured to: Acquire a text set corresponding to a specified user, and segment the text in the text set based on a comprehensive word library and a preset word segmentation algorithm to obtain a corresponding word segmentation result; wherein the text set includes all documents published by the specified user; Processing the word segmentation results by a preset statistical algorithm to obtain keywords in the text set; Determine a corresponding weight according to the specified keyword, and construct a keyword feature vector corresponding to the specified user according to the corresponding weight; Based on the pre-acquired keyword feature matrix and the preset filtering algorithm, determine the keywords recommended for the designated user according to the keyword feature vector corresponding to the designated user, and recommend the keywords to the designated user; The keyword feature matrix is composed of the keyword feature vectors corresponding to several users; Before determining the keywords recommended for the designated user, the method further includes: Obtain keyword feature vectors corresponding to a number of users from a pre-acquired keyword feature matrix; By using a preset filtering algorithm, respectively calculating the correlation coefficient distance between the keyword feature vector corresponding to the designated user and the keyword feature vectors corresponding to the plurality of users; Also includes: The correlation coefficient distance calculation formula between A and B is: Among them, A represents the specified user, B represents other users, and the keyword feature vector corresponding to the specified user A is X A =(x A1 ,x A2 ,…,x An ), the keyword feature vector corresponding to other user B is X B =(x B1 ,x B2 ,…,x Bn ); The step of determining the recommended keywords for the designated user based on the pre-acquired keyword feature matrix and the preset filtering algorithm according to the keyword feature vector corresponding to the designated user specifically includes: According to the correlation coefficient distance between the designated user and the plurality of users, a preset number of users whose correlation coefficient distance with the designated user is less than a preset distance threshold are obtained from the plurality of users; According to the correlation coefficient between the designated user and the plurality of users, weighted summing is performed on the keyword feature vectors corresponding to the preset number of users to obtain a corresponding keyword weight vector; Also includes: The formula for calculating keyword weight vector is: Among them, m is the preset number corresponding to the existing authors.
