Academic search sorting method, device, electronic device and storage medium

Abstract

The present invention provides an academic search sorting method, device, electronic device and storage medium, the method comprising: obtaining a search term input by a user; generating search results of the search term based on the text content of the search term and resource fusion data pre-stored according to multiple indexing methods; determining the priority sorting method of the search results based on the text length of the search term and multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function; obtaining the sorting result of the search results according to the relevance priority sorting method and the relevance between the search term and the search result. The method provided by the present invention can meet the search needs of different users, improve the accuracy and comprehensiveness of academic search, and enhance the user experience.

Description

Technical Field

[0001] The present invention relates to the field of information retrieval technology, and more specifically, to an academic search ranking method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of the Internet, Web services based on the HTTP protocol have become more and more popular, and the amount of resources and information on the Internet has increased dramatically. Users have the need to find resources distributed in various locations on the Internet based on their own personalized information.

[0003] Traditional academic search engines such as Google Scholar, Microsoft Academic Search, Baidu Academic Search, etc. have relatively complete functions for matching and finding relevant content in the document index library based on text content. However, since the index is mainly established based on the text relevance of the search content, the BM25 algorithm is a method used to measure the relevance between search engine queries and the queried text information. This algorithm is often used as an important feature in the calculation of search engine result ranking. However, the BM25 algorithm sorts the recall results based on the principle of short text retrieval priority. The sorting results of this algorithm cannot meet the needs of users for application or technology expansion retrieval based on a certain key technology or research field. Therefore, for some documents that are closely related to the search content but do not directly have a high text relevance, they are likely to be ignored in the BM25 algorithm mechanism, resulting in incomplete and inaccurate search results. Summary of the invention

[0004] The present invention provides an academic search ranking method, device, electronic device and storage medium, which are used to solve the technical problem that the short text search priority method used in the prior art causes incomplete and inaccurate search results, so as to achieve the purpose of improving the accuracy, comprehensiveness and personalized query of academic search.

[0005] In a first aspect, the present invention provides an academic search ranking method, comprising:

[0006] Get the search term entered by the user;

[0007] According to the text content of the search word, based on the resource fusion data pre-stored according to multiple indexing methods, obtaining the search results of the search word;

[0008] According to the text length of the search term, based on a plurality of relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function, determine a relevance priority sorting method corresponding to the search result;

[0009] Obtaining a ranking result of the search results according to the relevance priority ranking method and the relevance between the search term and the search result;

[0010] The resource fusion data refers to data obtained by associating data of different structural types.

[0011] In a second aspect, the present invention provides an academic search engine device, characterized in that it includes:

[0012] A first acquisition module, used to acquire a search term input by a user;

[0013] A first acquisition module, configured to acquire search results of the search term according to the text content of the search term and based on resource fusion data pre-stored according to multiple indexing methods;

[0014] A determination module, configured to determine, according to the text length of the search term, a relevance priority sorting method corresponding to the search result based on a plurality of relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function;

[0015] A third acquisition module, configured to acquire the ranking result of the search results according to the relevance priority ranking method and the relevance between the search term and the search result;

[0016] The resource fusion data refers to data obtained by associating data of different structural types.

[0017] In a third aspect, the present invention further provides an electronic device, comprising:

[0018] processor, memory and bus, wherein,

[0019] The processor and the memory communicate with each other via the bus;

[0020] The memory stores program instructions that can be executed by the processor, and the processor can execute any of the above methods by calling the program instructions.

[0021] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute any of the methods described above.

[0022] The present invention provides an academic search sorting method, device, electronic device and storage medium, the method comprising: obtaining a search term input by a user, obtaining the search results of the search term based on the text content of the search term and based on resource fusion data pre-stored according to multiple indexing methods, determining the relevance priority sorting method corresponding to the search results based on the text length of the search term and based on multiple relevance priority sorting methods pre-set according to the BM25 algorithm and Gaussian function, and obtaining the sorting result of the search results based on the relevance priority sorting method and the relevance between the search term and the search result. The method provided by the present invention can match the corresponding algorithm module with the search terms of different text lengths input by the user, obtain the corresponding search results, improve the accuracy and comprehensiveness of academic search, and enhance the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 A schematic diagram of a flow chart of an academic search ranking method provided by the present invention;

[0025] Figure 2 A schematic diagram of the overall process of academic search ranking based on the ABM25 algorithm provided by the present invention;

[0026] Figure 3 A schematic diagram of the structure of an academic search engine ranking device provided by the present invention;

[0027] Figure 4 This is a schematic structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0028] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 The flowchart of the academic search ranking method provided by the present invention is as follows. Figure 1 As shown, the academic search ranking method provided by the present invention comprises the following steps:

[0030] Step 101: Obtain the search term input by the user;

[0031] Step 102: According to the text content of the search word, based on the resource fusion data pre-stored according to multiple indexing methods, obtain the search result of the search word;

[0032] Step 103: According to the length of the search term, based on a plurality of relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function, obtaining a relevance priority sorting method corresponding to the search result;

[0033] Step 104: Obtaining a ranking result of the search results according to the relevance priority ranking method and the relevance between the search term and the search result;

[0034] The resource fusion data refers to data obtained by associating data of different structural types.

[0035] Specifically, the BM25 algorithm expands the scoring function of the binary independence model by adding document weights and query weights.

[0036] Among them, the academic search ranking method is implemented in an intelligent search engine, which refers to a search engine that can realize the reasoning and intelligent expansion of search terms, personalized sorting of search results, cross-resource graph recommendation, and resource relevance sorting. Among them, relevance refers to the degree of association between the search request and the search result.

[0037] In step 102, according to the text content of the output search term, the search result of the search term is obtained based on the pre-stored resource fusion data, wherein the resource fusion data stores data in the form of multiple indexes, and the multiple index forms may be full-text index, metadata index, and image index, etc., which are not specifically limited here.

[0038] In step 103, a plurality of different relevance priority ranking methods are obtained in advance in the intelligent search engine according to different parameters of the BM25 algorithm and the Gaussian function. The BM25 algorithm includes three free adjustment parameters, in addition to the adjustment factor b, there is also an adjustment factor k for word frequency. 1 and k 2, a variety of relevance priority sorting methods are obtained according to different parameter values. In this embodiment, the priority sorting method of the search results is selected according to the text length of the search term, and the priority sorting methods can be divided into short text relevance priority sorting method, long text relevance priority sorting method and mean text relevance priority sorting method. For example, when the text length of the search term falls within the set range of the short text relevance priority sorting method, the search results are sorted according to the short text relevance priority sorting method to obtain a sorting result. It should be noted that the setting of the priority sorting method can be set according to the actual needs of the user, and is not specifically limited here.

[0039] In step 104, the sorting result of the search results is obtained according to the relevance priority sorting method corresponding to the search results and the relevance between the search terms and the search results. For example, the search result sorting method determined by the search terms is a short text priority sorting method, and the search results are sorted according to the short text relevance priority sorting method, combined with the relevance between the search terms and the search results, to obtain the sorting result. It should be noted that in the embodiment of the present invention, the method of introducing a negative Gaussian function adjustment factor into the BM25 algorithm is called Academic BM25, abbreviated as ABM25, and in the intelligent search engine, it also includes some traditional relevance sorting algorithms, such as DFR, BM25, DFI and IB, among which DFR (Deviation frequency Random) refers to a random deviation algorithm, DFI (Deep / Dynamic Flow Inspection) refers to a deep / dynamic flow detection algorithm, and IB (Information Bottleneck) refers to an information bottleneck algorithm.

[0040] In an embodiment of the present invention, the search results of the search terms are obtained based on the text content of the search terms input by the user and based on resource fusion data stored in advance according to multiple indexing methods, and the priority sorting processing method corresponding to the search results is obtained based on the text length of the input search terms and the relevance priority sorting method pre-set based on the BM25 algorithm and different parameters of the Gaussian function. The sorting result is obtained according to the corresponding priority sorting processing method and the relevance between the search terms and the search results. The method provided by the present invention can match different relevance priority sorting methods according to the different text lengths of the search terms input by the user, and obtain different sorting methods of the search results, thereby meeting the search needs of different users, improving the accuracy and comprehensiveness of academic search, and enhancing user experience.

[0041] In another embodiment of the present invention, the method further comprises: generating a ranking result of personalized recommendation according to the personalized impact factor of the user based on the personalized recommendation mode of academic resources;

[0042] The personalized recommendation method for academic resources refers to a method for implementing personalized recommendations in the academic field based on the user's search habits.

[0043] Specifically, personalized ranking refers to the recommended ranking results obtained based on influencing factors such as each user's search interests, search fields, or long-term search habits.

[0044] In an embodiment of the present invention, the personalized recommendation method for academic resources is a method specifically used to realize personalized recommendations in the academic field based on the user's search habits. The personalized recommendation ranking results are obtained according to the user's personalized influence factors. It should be noted that the personalized influence factors can refer to the random walk model of the graph, the latent semantic model, the search behavior, the personal interest tags and other influence factors, and can also be other influence factors, which are not specifically limited here.

[0045] In an embodiment of the present invention, by introducing a personalized recommendation module for academic resources into an intelligent search engine, personalized recommendation ranking results are obtained based on personalized influencing factors such as the user's usual search habits, search behaviors, and personal interest tags, and personalized ranking result recommendations are implemented for users to review, thereby improving the user experience.

[0046] In another embodiment of the present invention, before generating the search results of the search term based on the resource fusion data pre-stored according to multiple indexing methods, the method further includes:

[0047] S1: Get resource data to be processed;

[0048] S2: extracting and performing attribute mapping processing on the resource data to be processed according to structured, semi-structured and unstructured data types to obtain first resource fusion data;

[0049] S3: performing quality assessment processing on the first resource fusion data according to the impact factor of the academic resource quality assessment to obtain second resource fusion data;

[0050] S4: Construct a full-text index, a metadata index, and a shadow image index based on the second resource fusion data.

[0051] Specifically, structured data refers to data that can be logically expressed and implemented by a two-dimensional table structure, strictly follows data format and length specifications, and is mainly stored and managed through relational databases; semi-structured data has a certain degree of structure; unstructured data is data with an irregular or incomplete structure, no predefined data model, and is not convenient to represent using a two-dimensional logical table in a database, including all formats of office documents, text, pictures, XML, HTML, various reports, images, and audio / video information.

[0052] In step S1, the resource data to be processed refers to academic literature, academicians and experts, patents, scientific and technological achievements and other resources in the academic field. It should be noted that the resource data to be processed can be set according to the actual needs of the user and is not specifically limited here.

[0053] In step S2, the resource data to be processed is extracted and attribute-mapped according to the three data types of structured, semi-structured and unstructured. That is to say, the various types of data are classified according to structured, semi-structured and unstructured, and the correlation between each other is established according to the attribute relationship between each data to achieve data fusion. Among them, attribute mapping refers to a processing method that establishes a corresponding relationship according to the attributes of the data, such as the title and the author. In an embodiment of the present invention, the first resource fusion data refers to the resource fusion data after extraction and attribute mapping. If an academic document is classified as structured data and a patent is classified as unstructured data, the correlation between the two authors is established according to the same attribute characteristics. When the corresponding academic document is retrieved, the patent also appears in the sorting result.

[0054] In step S3, the first resource fusion data is subjected to quality assessment processing according to the assessment factors of the academic resource quality assessment. The quality assessment is used to judge the integrity, standardization, consistency, accuracy, relevance, etc. of the resource fusion data, so that the assessed resource fusion data can be applied to the next step of index construction. In an embodiment of the present invention, the assessment factors of the academic resource quality assessment include compliance factors, expert influence factors, inclusion factors, etc., such as sorting all relevant resource data according to the level of expert influence, so that all resource data have a certain standardization. Among them, the assessment factors can be set according to specific needs and are not specifically limited here.

[0055] In step S4, a full-text index, a metadata index and a shadow image index are constructed based on the resource fusion data after quality assessment, wherein indexing refers to storing the processed resource fusion data in a database in a prescribed format so that the intelligent search engine can find the corresponding data when searching and sorting.

[0056] It should be noted that the principle of full-text indexing is to define a vocabulary, find the frequency and position of each term in the article, and summarize such frequency and position information in the order of the vocabulary, which is equivalent to establishing an index with the vocabulary as the directory for the file. When looking for a certain word, the position where the word appears can be quickly located; the metadata index is an index of the website site index compiled to facilitate users to search the site and improve the search speed, mainly including abstracts, invention names, author information, keywords and other content; the image index is an index established for information such as pictures in the resource data.

[0057] In an embodiment of the present invention, the resource data to be processed is extracted and attribute-mapped according to structured, semi-structured and unstructured, to obtain first resource fusion data, quality assessment is performed on the first resource fusion data according to the assessment factors of academic resource quality assessment, and a full-text index, metadata index and shadow image index are constructed based on the resource fusion data processed by the quality assessment. The method provided by the present invention can provide data support for intelligent engine search sorting and realize resource relevance sorting.

[0058] In another embodiment of the present invention, before extracting and mapping the attributes of the resource data according to structured, semi-structured and unstructured data types, the method includes:

[0059] Perform data cleaning processing on the resource data.

[0060] Specifically, data cleaning refers to the process of re-examining and verifying data, with the aim of removing duplicate information, correcting existing errors, and providing data consistency.

[0061] In the embodiment of the present invention, it is necessary to clean the resource data so that the cleaned data is more standardized and convenient for application in subsequent processing steps. It should be noted that data cleaning includes denoising, disambiguation, and normalization of resource data. Different cleaning steps can be selected according to actual needs, and are not specifically limited here.

[0062] In the embodiment of the present invention, by performing corresponding cleaning processing on the resource data, it can better meet the subsequent operation processing on the resource data and improve the accuracy of data classification.

[0063] In another embodiment of the present invention, the multiple relevance priority sorting methods include a short text relevance priority sorting method, a long text relevance priority sorting method and a mean text relevance priority sorting method. Accordingly, the relevance priority sorting method corresponding to the search result is determined based on the text length of the search term and the multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function, specifically including:

[0064] When the text length of the search term is less than or equal to a first preset threshold, determining that the relevance priority sorting method of the search results is the short text relevance priority sorting method;

[0065] or

[0066] When the text length of the search term is greater than the first preset threshold and less than the second preset threshold, determining that the relevance priority sorting method of the search results is the mean text relevance priority sorting method;

[0067] or

[0068] When the text length of the searched word is greater than or equal to the second preset threshold, it is determined that the relevance priority sorting method of the search results is the long text relevance priority sorting method.

[0069] Specifically, the first preset threshold and the second preset threshold are numerical values ​​representing the text length of the search term. The values ​​of the first preset threshold and the second preset threshold can be set according to actual needs and are not specifically limited here.

[0070] In an embodiment of the present invention, a short text relevance priority sorting method, a long text relevance priority sorting method or a mean text relevance priority sorting method is selected according to the text length of the search term. In this embodiment, the first preset threshold is set to 10 and the second preset threshold is set to 20. When the number of words in the search term is less than 10, the search results are sorted in a short text relevance priority sorting method; when the number of words in the search term is greater than 10 and less than 20, the search results are sorted in a mean text relevance priority sorting method to obtain a sorting result; when the number of words in the search term is greater than 20, the search results are sorted in a long text relevance priority sorting method to obtain a corresponding sorting result. It should be noted that the size of the first preset threshold and the second preset threshold can be set according to actual needs and is not specifically limited here.

[0071] In the embodiment of the present invention, the relevance priority sorting method of the search results is determined according to the text length of the search term, and the sorting result of the search results is obtained according to the determined relevance priority sorting method and the relevance between the search term and the search result. The present invention can meet the search needs of different users and improve the user experience.

[0072] In another embodiment of the present invention, the intelligent search engine refers to a search engine that can implement reasoning and intelligent expansion of search terms.

[0073] Specifically, extended search includes pinyin search and related search functions. For example, if the input text is "calculation", "computer", "calculation method", "calculation result", etc. may appear in the related search results, and multiple search results related to calculation may appear.

[0074] In the embodiment of the present invention, the intelligent search engine can realize the reasoning and intelligent expansion of search terms, and can meet the needs of realizing technical expansion retrieval according to the search terms input by the user.

[0075] Figure 3 A schematic diagram of the structure of an academic search engine ranking device provided by the present invention, such as Figure 3 As shown, the device provided by the present invention includes:

[0076] The first acquisition module 301 is used to acquire the search term input by the user;

[0077] A second acquisition module 302 is used to acquire search results of the search word according to the text content of the search word and based on resource fusion data pre-stored according to multiple indexing methods;

[0078] A determination module 303 is used to determine the relevance priority sorting method corresponding to the search result according to the text length of the search term and based on multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function;

[0079] A third acquisition module 304 is used to acquire the ranking result of the search result according to the relevance priority sorting method and the relevance between the search term and the search result;

[0080] The resource fusion data refers to data obtained by associating data of different structural types.

[0081] Specifically, an intelligent search engine refers to a search engine that can also perform reasoning and intelligent expansion of search terms.

[0082] In an embodiment of the present invention, a first acquisition module is used to acquire the text length of a search term input by a user; a second acquisition module is used to generate search results of the search term based on the text content of the search term and resource fusion data pre-stored according to a plurality of indexing methods; a determination module is used to determine a relevance priority sorting method corresponding to the search results according to a priority sorting method pre-set based on different parameters of a BM25 algorithm and a Gaussian function; a third acquisition module is used to obtain a sorting result according to a corresponding relevance priority sorting method and the relevance between the search term and the search result; the device provided by the present invention can match corresponding algorithm modules for search terms of different lengths input by users, and obtain corresponding search sorting results, which can meet the search needs of different users, improve the accuracy and comprehensiveness of academic search, and enhance user experience.

[0083] Since the principle of the device described in the embodiment of the present invention is the same as that of the method described in the above embodiment, a more detailed explanation will not be repeated here.

[0084] Figure 4 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 4 As shown, the present invention provides an electronic device, including: a processor (processor) 401, a memory (memory) 402 and a bus 403;

[0085] The processor 401 and the memory 402 communicate with each other via the bus 403;

[0086] The processor 401 is used to call the program instructions in the memory 402 to execute the methods provided by the above-mentioned method embodiments, for example, including: obtaining a search term input by a user, and obtaining the search results of the search term based on the text content of the search term and based on resource fusion data pre-stored according to multiple indexing methods; determining the relevance priority sorting method corresponding to the search results based on the text length of the search term and based on multiple relevance priority sorting methods pre-set according to the BM25 algorithm and Gaussian function; obtaining the sorting result of the search results based on the relevance priority sorting method and the relevance between the search term and the search result; wherein the resource fusion data refers to data obtained by associating data of different structural types.

[0087] The present embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable the computer to execute the methods provided by the above-mentioned method embodiments, for example, including: obtaining a search term input by a user, and obtaining search results of the search term based on the text content of the search term and resource fusion data pre-stored according to multiple indexing methods; determining a relevance priority sorting method corresponding to the search result based on a plurality of relevance priority sorting methods pre-set according to a BM25 algorithm and a Gaussian function according to the text length of the search term; obtaining a sorting result of the search result based on the relevance priority sorting method and the relevance between the search term and the search result; wherein the resource fusion data refers to data obtained by associating data of different structural types.

[0088] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for ranking academic searches, characterized in that: include: Get the search term entered by the user; According to the text content of the search word, based on the resource fusion data pre-stored according to multiple indexing methods, obtaining the search results of the search word; According to the text length of the search term, based on a plurality of relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function, determine a relevance priority sorting method corresponding to the search result; Obtaining a ranking result of the search results according to the relevance priority ranking method and the relevance between the search term and the search result; The resource fusion data refers to data obtained by associating data of different structural types; The multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function include multiple different relevance priority sorting methods pre-obtained in the intelligent search engine according to different parameters of the BM25 algorithm and the Gaussian function; wherein the BM25 algorithm contains free adjustment parameters, the free adjustment parameters include the adjustment factor b and the adjustment factors k1 and k2 for the word frequency, and a negative Gaussian function adjustment factor is introduced into the BM25 algorithm; multiple relevance priority sorting methods are obtained according to different parameter values ​​of the BM25 algorithm and the Gaussian function.

2. The academic search ranking method according to claim 1, characterized in that: The method further comprises: Based on the personalized recommendation method of academic resources, generating a ranking result of personalized recommendation of search results according to the personalized influence factor of the user; The personalized recommendation method for academic resources refers to a method for implementing personalized recommendations in the academic field based on the user's search habits.

3. The academic search ranking method according to claim 1, characterized in that: Before generating the search results of the search terms based on the resource fusion data pre-stored according to multiple indexing methods, the method further includes: S1: Get resource data to be processed; S2: extracting and performing attribute mapping processing on the resource data to be processed according to structured, semi-structured and unstructured data types to obtain first resource fusion data; S3: performing quality assessment processing on the first resource fusion data according to the assessment factors of academic resource quality assessment to obtain second resource fusion data; S4: Construct a full-text index, a metadata index, and a shadow image index based on the second resource fusion data.

4. The academic search ranking method according to claim 3, characterized in that: Before extracting and attribute mapping the resource data according to structured, semi-structured and unstructured data types, the method includes: Perform data cleaning processing on the resource data.

5. The academic search ranking method according to claim 1, characterized in that: The multiple relevance priority sorting methods include a short text relevance priority sorting method, a long text relevance priority sorting method and a mean text relevance priority sorting method. Accordingly, the relevance priority sorting method corresponding to the search result is determined based on the text length of the search term and the multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function, specifically including: When the text length of the search term is less than or equal to a first preset threshold, determining that the relevance priority sorting method of the search results is the short text relevance priority sorting method; or When the text length of the search term is greater than the first preset threshold and less than the second preset threshold, determining that the relevance priority sorting method of the search results is the mean text relevance priority sorting method; or When the text length of the search term is greater than or equal to the second preset threshold, it is determined that the relevance priority sorting method of the search results is the long text relevance priority sorting method.

6. An academic search ranking device, characterized in that: include: A first acquisition module, used to acquire a search term input by a user; A second acquisition module is used to acquire search results of the search word according to the text content of the search word and based on resource fusion data pre-stored according to multiple indexing methods; A determination module, configured to determine, according to the text length of the search term, a relevance priority sorting method corresponding to the search result based on a plurality of relevance priority sorting methods pre-set according to different parameters of the BM25 algorithm and the Gaussian function; A third acquisition module, configured to acquire the ranking result of the search results according to the relevance priority ranking method and the relevance between the search term and the search result; The resource fusion data refers to data obtained by associating data of different structural types; The multiple relevance priority sorting methods pre-set according to the BM25 algorithm and the Gaussian function include multiple different relevance priority sorting methods pre-obtained in the intelligent search engine according to different parameters of the BM25 algorithm and the Gaussian function; wherein the BM25 algorithm contains free adjustment parameters, the free adjustment parameters include the adjustment factor b and the adjustment factors k1 and k2 for the word frequency, and a negative Gaussian function adjustment factor is introduced into the BM25 algorithm; multiple relevance priority sorting methods are obtained according to different parameter values ​​of the BM25 algorithm and the Gaussian function.

7. The academic search ranking device according to claim 6, characterized in that: The device also includes: A generation module generates a ranking result of personalized recommendation according to the personalized influence factor of the user.

8. The academic search ranking device according to claim 6, characterized in that: The apparatus further comprises a building block for: Get the resource data to be processed; Extracting and attribute mapping the resource data to be processed according to structured, semi-structured and unstructured data types to obtain first resource fusion data; According to the evaluation factors of the academic resource quality evaluation, the resource fusion data is subjected to quality evaluation processing to obtain second resource fusion data; According to the second resource fusion data, a full-text index, a metadata index and a shadow image index are constructed.

9. An electronic device, characterized in that: include: processor, memory and bus, wherein, The processor and the memory communicate with each other via the bus; The memory stores program instructions that can be executed by the processor, and the processor can execute any one of the methods according to claims 1 to 5 by calling the program instructions.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, which cause the computer to execute the method according to any one of claims 1 to 5.

Images