Method and device for optimizing quotation linking speed, storage medium and computer equipment

Abstract

The invention discloses a citation linking speed optimization method and device, a storage medium and computer equipment, relates to the technical field of data retrieval, and mainly aims at solving the problem that a retrieval engine optimization scheme adopted during existing citation linking lacks data optimization consideration. Dividing the literature bibliography data pool based on a multi-level division rule to obtain a plurality of sub-literature bibliography data pools; extracting keywords in the to-be-linked citation information based on a multi-level division rule, and performing classification processing on the to-be-linked citation information based on the keywords to obtain a target citation information category; obtaining an association relationship between the citation information category and a sub-literature bibliography data pool, and determining a target sub-literature bibliography data pool corresponding to the target citation information category based on the association relationship; and retrieving target literature information corresponding to the to-be-linked quotation information from the target sub-literature bibliography data pool, and establishing a literature link based on the target literature information.

Description

Technical Field

[0001] The present invention relates to the technical field of data retrieval, and in particular to a method and device for optimizing citation link speed, a storage medium, and a computer device. Background Art

[0002] Citation linking is the process of finding the corresponding literature in the subject database by analyzing and calculating the character information of the references written by the author of the article. With the development of Internet technology and the progress of domestic science and technology, the growth of domestic scientific and technological literature has increased year by year. This has led to two problems: first, there are more and more citations to be linked; second, the subject database to be searched is getting larger and larger. In actual data calculation, the retrieval time of citation links is getting longer and longer. At present, the method of optimizing the search engine is used to retrieve this massive level of data, such as using multi-threaded concurrency and other technologies to improve efficiency. However, the existing citation linking scheme has not found a way to optimize efficiency based on the characteristics of the data itself. Summary of the invention

[0003] In view of this, the present invention provides a method and device for optimizing citation link speed, a storage medium, and a computer device, the main purpose of which is to solve the problem that the existing search engine optimization scheme used in citation linking lacks optimization considerations for the characteristics of the data itself.

[0004] According to one aspect of the present invention, a method for optimizing citation link speed is provided, comprising:

[0005] Obtaining a preset multi-level division rule, and dividing the document title data pool based on the multi-level division rule to obtain a plurality of sub-document title data pools;

[0006] Extracting keywords from the citation information to be linked based on the multi-level division rule, and classifying the citation information to be linked based on the keywords to obtain a target citation information category;

[0007] Acquire the association relationship between the citation information category and the sub-document bibliographic data pool, and determine the target sub-document bibliographic data pool corresponding to the target citation information category based on the association relationship;

[0008] Target document information corresponding to the citation information to be linked is retrieved from the target sub-document title data pool, and a document link is established based on the target document information.

[0009] Furthermore, the document title data pool is divided based on the multi-level division rule to obtain multiple sub-document title data pools, including:

[0010] Acquire the hierarchical features corresponding to each level in the multi-level division rule; and combine the hierarchical features to obtain document feature information corresponding to each sub-document title data pool;

[0011] The document information in the document bibliographic data pool is screened and processed in sequence according to the hierarchical order and based on the hierarchical features, so as to obtain the sub-document bibliographic data pool corresponding to each document characteristic information.

[0012] Furthermore, the screening of the document information in the document bibliographic data pool in sequence according to the hierarchical order based on the hierarchical features includes:

[0013] Acquire first-level features, and screen the document information in the document bibliographic data pool based on the first-level features to obtain first-level screening results;

[0014] Acquire second-level features, and perform screening processing on the document information in the first-level screening results based on the second-level features to obtain second-level screening results;

[0015] A third-level feature is obtained, and based on the third-level feature, the document information in the second-level screening result is screened to obtain a third-level screening result.

[0016] Furthermore, extracting keywords from the citation information to be linked based on the multi-level division rule includes:

[0017] Identifying the separation points in the citation information to be linked, and performing block processing on the citation information to be linked based on the separation points to obtain a block result of the citation information to be linked;

[0018] Obtaining hierarchical features corresponding to each level in the multi-level division rule;

[0019] The keyword is extracted from the to-be-linked citation block result based on the hierarchical feature.

[0020] Furthermore, before classifying the citation information to be linked based on the keyword to obtain the target citation information category, the method further includes:

[0021] A plurality of citation information categories equal in number to the sub-document title data pools are set; the citation information categories have the same document characteristic information as the sub-document title data pools;

[0022] Acquire a citation information sample data set, and perform block processing on the citation information samples in the citation information sample data set to obtain a block result;

[0023] Keyword information is extracted from the segmentation results, and a citation classification model is trained based on the keyword information to obtain a citation classification model for identifying the category of the citation information.

[0024] Furthermore, before obtaining the association between the citation information category and the sub-document bibliographic data pool, the method further includes:

[0025] Respectively obtain document feature information of the sub-document title data pool; and obtain citation information categories having the same document feature information;

[0026] The document feature information is associated with the document feature category to obtain an association relationship between the citation information category and the sub-document bibliographic data pool.

[0027] Furthermore, the method further comprises:

[0028] Obtain new document data at preset time intervals;

[0029] Based on the multi-level division rule, the newly added document data is divided and processed to obtain a newly added document division result;

[0030] Based on the newly added document classification result, data updating processing is performed on each of the sub-document title data pools to obtain an updated sub-document title data pool.

[0031] According to another aspect of the present invention, there is provided a device for optimizing citation link speed, comprising:

[0032] A data pool partitioning module is used to obtain a preset multi-level partitioning rule, and partition the document title data pool based on the multi-level partitioning rule to obtain multiple sub-document title data pools;

[0033] A citation classification module, used for extracting keywords from the citation information to be linked based on the multi-level division rule, and classifying the citation information to be linked based on the keywords to obtain a target citation information category;

[0034] A target determination module, used to obtain the association relationship between the citation information category and the sub-document bibliographic data pool, and determine the target sub-document bibliographic data pool corresponding to the target citation information category based on the association relationship;

[0035] The retrieval module is used to retrieve the target document information corresponding to the citation information to be linked from the target sub-document title data pool, and establish a document link based on the target document information.

[0036] Furthermore, the data pool partitioning module includes:

[0037] A document feature extraction unit is used to obtain the hierarchical features corresponding to each level in the multi-level division rule; and to combine the hierarchical features to obtain document feature information corresponding to each sub-document title data pool;

[0038] The screening and dividing unit is used to screen the document information in the document title data pool in sequence according to the hierarchical order based on the hierarchical characteristics to obtain the sub-document title data pool corresponding to each of the document characteristic information.

[0039] Furthermore, the screening and dividing unit is also used for:

[0040] Acquire first-level features, and screen the document information in the document bibliographic data pool based on the first-level features to obtain first-level screening results;

[0041] Acquire second-level features, and perform screening processing on the document information in the first-level screening results based on the second-level features to obtain second-level screening results;

[0042] A third-level feature is obtained, and based on the third-level feature, the document information in the second-level screening result is screened to obtain a third-level screening result.

[0043] Furthermore, the citation classification module is also used for:

[0044] Identifying the separation points in the citation information to be linked, and performing block processing on the citation information to be linked based on the separation points to obtain a block result of the citation information to be linked;

[0045] Obtaining hierarchical features corresponding to each level in the multi-level division rule;

[0046] The keyword is extracted from the to-be-linked citation block result based on the hierarchical feature.

[0047] Furthermore, the device also includes a classification model training module, which is used to:

[0048] A plurality of citation information categories equal in number to the sub-document title data pools are set; the citation information categories have the same document characteristic information as the sub-document title data pools;

[0049] Acquire a citation information sample data set, and perform block processing on the citation information samples in the citation information sample data set to obtain a block result;

[0050] Keyword information is extracted from the segmentation results, and a citation classification model is trained based on the keyword information to obtain a citation classification model for identifying the category of the citation information.

[0051] Furthermore, the device also includes an association module, which is used to:

[0052] Respectively obtain document feature information of the sub-document title data pool; and obtain citation information categories having the same document feature information;

[0053] The document feature information is associated with the document feature category to obtain an association relationship between the citation information category and the sub-document bibliographic data pool.

[0054] Furthermore, the device also includes a data updating module, which is used to:

[0055] Obtain new document data at preset time intervals;

[0056] Based on the multi-level division rule, the newly added document data is divided and processed to obtain a newly added document division result;

[0057] Based on the newly added document classification result, data updating processing is performed on each of the sub-document title data pools to obtain an updated sub-document title data pool.

[0058] According to another aspect of the present invention, a storage medium is provided, wherein the storage medium stores at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the above-mentioned method for optimizing citation link speed.

[0059] According to another aspect of the present invention, there is provided a computer device, comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0060] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for optimizing the citation link speed.

[0061] By means of the above technical solution, the technical solution provided by the embodiment of the present invention has at least the following advantages:

[0062] The present invention provides a method and device for optimizing the citation link speed, a storage medium, and a computer device. Compared with the prior art, the present invention can extract features of different types of data according to different needs by presetting a multi-level division rule with high flexibility. The document title data pool is divided and processed by the multi-level division rule to obtain multiple sub-document title data pools; and the keywords in the citation information to be linked are extracted by the multi-level division rule, and the citation information to be linked is classified and processed based on the keywords to obtain the target citation information category. Then, the association relationship between the citation information category and the sub-document title data pool is obtained, and the target sub-document title data pool corresponding to the target citation information category is determined, so as to optimize the retrieval speed based on the characteristics of the data itself. The present invention reduces the amount of data scanning and improves the single retrieval speed, and finally accelerates the citation link speed, and shares the growing data pressure to each sub-data pool, which greatly alleviates the growing data pressure.

[0063] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0065] Figure 1 A schematic flow chart of a method for optimizing citation link speed provided by an embodiment of the present invention is shown;

[0066] Figure 2 A schematic diagram of a multi-level division rule provided by an embodiment of the present invention is shown;

[0067] Figure 3 The diagram shows an association relationship between a citation information category and a sub-document title data pool provided by an embodiment of the present invention;

[0068] Figure 4 A schematic flow chart of another method for optimizing citation link speed provided by an embodiment of the present invention is shown;

[0069] Figure 5 A schematic diagram of document feature information provided by an embodiment of the present invention is shown;

[0070] Figure 6A schematic flow chart of another method for optimizing citation link speed provided by an embodiment of the present invention is shown;

[0071] Figure 7 A schematic flow chart of another method for optimizing citation link speed provided by an embodiment of the present invention is shown;

[0072] Figure 8 A schematic diagram showing the structure of another device for optimizing citation link speed provided by an embodiment of the present invention is shown;

[0073] Fig. 9 A schematic structural diagram of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0074] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0075] At present, the usual process of citation linking is: first, two data pools are formed, one is the data pool of citation information to be linked, and the other is the data pool of retrieved document titles. In actual situations, the data pool of citation information to be linked has about 1.4 billion records, and the data pool of retrieved document titles has about 630 million records. Then extract the data to be linked one by one, combine multiple search statements according to the citation characters, and search in the data pool of retrieved document titles. Finally, find the most likely consistent data through calculation and judgment. If there are m pieces of citation information to be linked and n pieces of document titles to be retrieved, then m*n pieces of data need to be involved in the retrieval to complete all citation links. For citation links with massive data scale, this number is extremely large.

[0076] The embodiment of the present invention provides a method for optimizing the citation link speed, such as Figure 1 As shown, the method includes:

[0077] 101. Obtain a preset multi-level division rule, and divide the document title data pool based on the multi-level division rule to obtain multiple sub-document title data pools;

[0078] In an embodiment of the present invention, the current execution end obtains a preset multi-level division rule, wherein the multi-level division rule can be set according to the division needs. For example, the first level can select the simplest feature for division, such as being divided into different types of languages ​​such as Chinese and English. Of course, each type of feature at this level still contains a large amount of data, and it is impossible to achieve the original design intention of greatly reducing the amount of retrieval scanning; therefore, features can continue to be extracted at the second level, such as the document type of a single citation information can be proposed, that is, which type of document it belongs to, including journals, conference proceedings, doctoral and master's theses, books, newspapers, etc. At this level, the classification features are preliminarily refined, but for document types with large publication volumes, such as journals, doctoral and master's theses, etc., there are still tens of millions or hundreds of millions of data. Therefore, a third level can also be set for division, based on the second subdivision of the document type, such as, documents of the journal type can also be divided by journal name; documents of the conference proceedings type can also be divided by conference name; documents of the doctoral and master's thesis type can also be divided by school, etc. For details, please refer to Figure 2 In this embodiment, the multi-level division rules can also be adjusted as needed, such as adjusting the number of division levels, and adjusting the division types of each level, etc., which are not specifically limited in this embodiment of the present invention.

[0079] In the embodiment of the present invention, after obtaining the multi-level division rule, the current execution end also divides the document title data pool based on the multi-level division rule to obtain multiple sub-document title data pools. Figure 2 As shown, the first level is divided by language type to obtain 2 data pools, and the second level is divided by document type, and the data pool corresponding to the Chinese type is divided into 6 data pools, and the data pool corresponding to the corresponding foreign language type can also be divided into multiple data pools. Therefore, after multi-level division, multiple sub-document title data pools can be obtained, which is not specifically limited in the embodiment of the present invention.

[0080] 102. Extract keywords from the citation information to be linked based on the multi-level division rule, and classify the citation information to be linked based on the keywords to obtain a target citation information category;

[0081] In an embodiment of the present invention, the current execution end extracts keywords from the citation information to be linked based on a multi-level division rule. The purpose of keyword extraction is to classify the citation information to be linked according to the multi-level division rule obtained in step 101, so as to determine the retrieval target of the citation information to be linked in the sub-bibliographic data pool obtained in step 101. For example, when the citation information to be linked is "He Moumou (2023), Research on the Construction of Female Images in Domestic Online Self-produced Talk Shows [D] Doctoral Dissertation. Yinchuan: Ningxia University", keywords such as "D", "Doctoral Dissertation", and "Ningxia University" can be extracted as the keywords of the citation information to be linked, etc., and the embodiments of the present invention do not make specific limitations. Based on the multi-level division rule exemplified in step 101, the above citation information to be linked can be classified into the target citation information category such as "Chinese Doctoral and Master's Theses, Ningxia University", etc., and the embodiments of the present invention do not make specific limitations.

[0082] 103. Obtain the association relationship between the citation information category and the sub-bibliographic data pool, and determine the target sub-bibliographic data pool corresponding to the target citation information category based on the association relationship;

[0083] In an embodiment of the present invention, the current execution end obtains the association relationship between the citation information category and the sub-bibliographic data pool. The association relationship is preset according to the multi-level division rule and is used to represent the target retrieval data pool corresponding to each citation information category, which can greatly reduce the data scanning volume. For example Figure 3 As shown, the data scanning volume of the prior art is m*n, that is, any citation information to be linked needs to scan n records in the bibliographic data pool. In the present invention, the citation information to be linked only needs to scan the target sub-bibliographic data pool corresponding to the target citation information category to accurately retrieve the target document. Figure 3 Among them, the total scanning volume of m citation information to be linked changes from m*n to m1*n1 + m2*n2 +... m i *n i (where m1 + m2 + m i = m, n1 + n2 + n i = n), from which it can be seen that m1*n1 + m2*n2 +... m i *n i < m*n, and for a huge amount of data in the hundreds of millions, the fewer the records in the retrieved data pool, the faster the retrieval speed.

[0084] 104. Retrieve the target document information corresponding to the citation information to be linked from the target sub-bibliographic data pool, and establish a document link based on the target document information.

[0085] In the embodiment of the present invention, the current execution end retrieves the target document information corresponding to the citation information to be linked from the target sub-document title data pool determined in step 103, and then establishes a document link based on the target document information.

[0086] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to accurately divide the literature bibliographic data pool and obtain a division result consistent with the multi-level division rule, another method for optimizing the citation link speed is provided, such as Figure 4 As shown, the step divides the document title data pool based on the multi-level division rule to obtain multiple sub-document title data pools, including:

[0087] 201. Obtaining hierarchical features corresponding to each hierarchical level in the multi-level division rule; and combining the hierarchical features to obtain document feature information corresponding to each sub-document title data pool;

[0088] In the embodiment of the present invention, the current execution end obtains the hierarchical features corresponding to each level in the multi-level division rule, such as Figure 2 As shown, the hierarchical features of the first level are "Chinese" and "foreign"; the hierarchical features of the second level include "journals", "books", "doctoral and master's thesis", "conference proceedings", etc.; the hierarchical features of the third level include "journal name", "conference name", "school", etc., which are not specifically limited in the embodiment of the present invention. After obtaining the hierarchical features, the current execution end combines the hierarchical features to obtain the document feature information corresponding to each sub-document title data pool. Figure 5 As shown, the combined "Chinese Journal Name 1", "Chinese Journal Name 2", "Chinese Book Publishing House 1", "Chinese Book Publishing House 2", etc. are document feature information corresponding to each sub-document title data pool, and the embodiment of the present invention does not make specific limitations.

[0089] 202. Screening and processing the document information in the document title data pool in sequence according to the hierarchical order and based on the hierarchical characteristics, to obtain the sub-document title data pool corresponding to each of the document characteristic information.

[0090] In the embodiment of the present invention, the current execution end sequentially screens and processes the document information in the document title data pool based on the hierarchical characteristics in the order of the hierarchical levels to obtain a sub-document title data pool corresponding to each document feature information. Since the document title data pool at the retrieved end is structured data, it can be easily divided into tables according to the document feature information through structured fields.

[0091] It should be noted that the steps of screening the document information in the document bibliographic data pool in sequence according to the hierarchical order based on the hierarchical features include:

[0092] The first-level features are obtained, and the document information in the document bibliographic data pool is screened based on the first-level features to obtain the first-level screening results. For example, the first-level screening results corresponding to the "Chinese" feature and the first-level screening results corresponding to the "foreign" feature are obtained, which are not specifically limited in the embodiments of the present invention.

[0093] Acquire the second-level features, and filter and process the document information in the first-level screening results based on the second-level features to obtain the second-level screening results. For example, further filter the first-level screening results corresponding to the above-mentioned "Chinese" feature to obtain the second-level screening results corresponding to the "Chinese journal" feature, the second-level screening results corresponding to the "Chinese book" feature, etc., which are not specifically limited in the embodiments of the present invention. Similarly, further filter the first-level screening results corresponding to the above-mentioned "foreign language" feature to obtain the second-level screening results corresponding to the "foreign language book" feature, the second-level screening results corresponding to the "foreign language patent" feature, etc., which are not specifically limited in the embodiments of the present invention.

[0094] The third-level feature is obtained, and the document information in the second-level screening result is screened and processed based on the third-level feature to obtain the third-level screening result. For example, the second-level screening result corresponding to the above-mentioned "Chinese journal" feature is further screened to obtain the third-level screening result corresponding to the "Chinese journal electromechanical engineering" feature, the third-level screening result corresponding to the "Chinese journal mechanical transmission" feature, etc., which is not specifically limited in the embodiment of the present invention.

[0095] It should be noted that, since the multi-level division rule can also adjust the number of division levels and / or the division type of each level as needed, the process of sequentially screening and processing the document information in the document title data pool according to the order of the levels based on the level characteristics in the embodiment of the present invention is not limited to the above-mentioned example description. For example, the division and processing according to the year of publication or disclosure can also be set after the third level, and the embodiment of the present invention does not make specific limitations.

[0096] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to more accurately obtain the feature information involved in the multi-level division rule so as to accurately classify the citation information to be linked, another method for optimizing the citation link speed is provided, such as Figure 6 As shown, the step of extracting keywords in the citation information to be linked based on the multi-level division rule includes:

[0097] 301. Identify the separation points in the citation information to be linked, and perform block processing on the citation information to be linked based on the separation points to obtain a block result of the citation information to be linked;

[0098] 302. Obtaining hierarchical features corresponding to each hierarchical level in the multi-level division rule;

[0099] 303. Extract the keyword from the to-be-linked citation block results based on the hierarchical features.

[0100] In an embodiment of the present invention, the current execution end identifies the separation point in the citation information to be linked, such as identifying symbols such as commas, dots, colons, spaces, etc. as separation points, which are not specifically limited in the embodiment of the present invention. After identifying the separation point, the current execution end performs block processing on the citation information to be linked based on the separation point to obtain the block result of the citation to be linked; then, the hierarchical features corresponding to each level in the multi-level division rule are obtained in the manner shown in steps 201 and 202; then, keywords are extracted from the block result of the citation to be linked based on the hierarchical features. For example, when the citation information to be linked is "He Moumou (2023), Doctoral Dissertation on the Construction of Female Images in Domestic Online Self-Made Talk Shows [D]. Yinchuan: Ningxia University", "D", "Doctoral Dissertation", "Ningxia University", etc. can be extracted as keywords of the citation information to be linked, which are not specifically limited in the embodiment of the present invention.

[0101] It should be noted that the embodiments of the present invention require machine learning, that is, the use of deep learning technology to block and classify the citation information to achieve the extraction of citation information. This task uses a convolutional neural network combined with Bi-GRU and CRF algorithms to build a citation block model, block the original citation information, and then use the deep learning pre-training model Bert for semantic understanding combined with the attention mechanism and CRF algorithm to build a citation classification model to classify and identify the block information. The extraction of citation information is achieved by connecting the above two models, and the embodiments of the present invention do not make specific limitations.

[0102] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to improve the accuracy of document retrieval, another method for optimizing the citation link speed is provided, such as Figure 7 As shown, before the step of classifying the citation information to be linked based on the keyword to obtain the target citation information category, the method further includes:

[0103] 401. Setting a plurality of citation information categories equal to the number of the sub-document title data pools; the citation information categories and the sub-document title data pools have the same document characteristic information;

[0104] In an embodiment of the present invention, the current execution end sets a plurality of citation information categories equal to the number of the sub-document title data pool; since the citation information categories and the sub-document title data pool are both obtained based on a preset multi-level division rule, the citation information categories in the embodiment of the present invention have the same document characteristic information as the sub-document title data pools divided from step 201 to step 202.

[0105] 402. Acquire a citation information sample data set, and perform block processing on the citation information samples in the citation information sample data set to obtain a block result;

[0106] 403. Extract keyword information from the segmentation results, and train a citation classification model based on the keyword information to obtain a citation classification model for identifying the category of the citation information.

[0107] In the embodiment of the present invention, the current execution end obtains a citation information sample data set, and uses the same technical means as step 301 to perform block processing on the citation information samples in the citation information sample data set to obtain a block result. Then, the same technical means as step 303 are used to extract keyword information from the block result. Next, the citation classification model is trained based on the extracted keyword information to obtain a citation classification model for identifying citation information categories.

[0108] It should be noted that the embodiments of the present invention require machine learning, that is, the use of deep learning technology to block and classify the citation information to achieve the extraction of citation information. This task uses a convolutional neural network combined with Bi-GRU and CRF algorithms to build a citation block model, block the original citation information, and then use the deep learning pre-training model Bert for semantic understanding combined with the attention mechanism and CRF algorithm to build a citation classification model to classify and identify the block information. The extraction of citation information is achieved by connecting the above two models, and the embodiments of the present invention do not make specific limitations.

[0109] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to more quickly determine the target sub-document bibliographic data pool, another method for optimizing the citation link speed is provided. Before obtaining the association relationship between the citation information category and the sub-document bibliographic data pool, the method further includes:

[0110] Respectively obtain document feature information of the sub-document title data pool; and obtain citation information categories having the same document feature information;

[0111] The document feature information is associated with the document feature category to obtain an association relationship between the citation information category and the sub-document bibliographic data pool.

[0112] In the embodiment of the present invention, the current execution end respectively obtains the document feature information of the sub-document title data pool obtained in step 201; and obtains the citation information category with the same document feature information based on step 401; then, the obtained document feature information is associated with the citation information category with the same document feature information, and the association relationship between the citation information category and the sub-document title data pool is obtained, such as Figure 3 As shown, the citation information category m1 is associated with the sub-document title data pool n1; the citation information category m2 is associated with the sub-document title data pool n2, ..., the citation information category m i and sub-bibliographic data pool n i The embodiments of the present invention do not make any specific limitations on the above.

[0113] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to adapt to the increasingly large database of retrieved document titles, another method for optimizing the citation link speed is provided, and the method further includes:

[0114] Obtain new document data at preset time intervals;

[0115] Based on the multi-level division rule, the newly added document data is divided and processed to obtain a newly added document division result;

[0116] Based on the newly added document classification result, data updating processing is performed on each of the sub-document title data pools to obtain an updated sub-document title data pool.

[0117] In an embodiment of the present invention, the current execution end obtains the newly added document data according to a preset time interval, wherein the time interval can be set to 24 hours, 48 ​​hours, etc., which is not specifically limited in the embodiment of the present invention. Then, the newly added document data is divided and processed based on the multi-level division rule in step 101 to obtain the newly added document division result. That is, the newly added documents are divided into each sub-document title data pool according to the original multi-level division rule. The current execution end performs data update processing on each sub-document title data pool based on the newly added document division result to obtain an updated sub-document title data pool.

[0118] It should be noted that when the amount of data in the sub-document title data pool reaches a certain level, the atomic document title data pool can be further divided as needed to obtain a document title data pool with a smaller amount of data. In addition, in order to adapt to multi-threaded concurrent technology, all the citation information to be linked can be first extracted through feature extraction to form a set, and each set can be retrieved from the corresponding sub-document title data pool, which is not specifically limited in the embodiment of the present invention.

[0119] The embodiment of the present invention provides a method for optimizing the citation link speed. Compared with the prior art, the present invention can extract features of different types of data according to different needs by presetting a multi-level division rule with high flexibility. The document title data pool is divided and processed by the multi-level division rule to obtain multiple sub-document title data pools; and the keywords in the citation information to be linked are extracted by the multi-level division rule, and the citation information to be linked is classified and processed based on the keywords to obtain the target citation information category. Then, the association relationship between the citation information category and the sub-document title data pool is obtained, and the target sub-document title data pool corresponding to the target citation information category is determined, so as to optimize the retrieval speed based on the characteristics of the data itself. The present invention reduces the amount of data scanning and improves the single retrieval speed, and finally accelerates the citation link speed, and shares the growing data pressure to each sub-data pool, which greatly alleviates the growing data pressure.

[0120] As the above Figure 1 The embodiment of the present invention provides a device for optimizing the citation link speed, such as Figure 8 As shown, the device comprises:

[0121] The data pool partitioning module 51 is used to obtain a preset multi-level partitioning rule, and partition the document title data pool based on the multi-level partitioning rule to obtain a plurality of sub-document title data pools;

[0122] A citation classification module 52, for extracting keywords from the citation information to be linked based on the multi-level classification rule, and classifying the citation information to be linked based on the keywords to obtain a target citation information category;

[0123] A target determination module 53 is used to obtain the association relationship between the citation information category and the sub-document title data pool, and determine the target sub-document title data pool corresponding to the target citation information category based on the association relationship;

[0124] The retrieval module 54 is used to retrieve the target document information corresponding to the citation information to be linked from the target sub-document title data pool, and establish a document link based on the target document information.

[0125] Furthermore, the data pool partitioning module 51 includes:

[0126] A document feature extraction unit is used to obtain the hierarchical features corresponding to each level in the multi-level division rule; and to combine the hierarchical features to obtain document feature information corresponding to each sub-document title data pool;

[0127] The screening and dividing unit is used to screen the document information in the document title data pool in sequence according to the hierarchical order based on the hierarchical characteristics to obtain the sub-document title data pool corresponding to each of the document characteristic information.

[0128] Furthermore, the screening and dividing unit is also used for:

[0129] Acquire first-level features, and screen the document information in the document bibliographic data pool based on the first-level features to obtain first-level screening results;

[0130] Acquire second-level features, and perform screening processing on the document information in the first-level screening results based on the second-level features to obtain second-level screening results;

[0131] A third-level feature is obtained, and based on the third-level feature, the document information in the second-level screening result is screened to obtain a third-level screening result.

[0132] Furthermore, the citation classification module 52 is also used for:

[0133] Identifying the separation points in the citation information to be linked, and performing block processing on the citation information to be linked based on the separation points to obtain a block result of the citation information to be linked;

[0134] Obtaining hierarchical features corresponding to each level in the multi-level division rule;

[0135] The keyword is extracted from the to-be-linked citation block result based on the hierarchical feature.

[0136] Furthermore, the device also includes a classification model training module, which is used to:

[0137] A plurality of citation information categories equal in number to the sub-document title data pools are set; the citation information categories have the same document characteristic information as the sub-document title data pools;

[0138] Acquire a citation information sample data set, and perform block processing on the citation information samples in the citation information sample data set to obtain a block result;

[0139] Keyword information is extracted from the segmentation results, and a citation classification model is trained based on the keyword information to obtain a citation classification model for identifying the category of the citation information.

[0140] Furthermore, the device also includes an association module, which is used to:

[0141] Respectively obtain document feature information of the sub-document title data pool; and obtain citation information categories having the same document feature information;

[0142] The document feature information is associated with the document feature category to obtain an association relationship between the citation information category and the sub-document bibliographic data pool.

[0143] Furthermore, the device also includes a data updating module, which is used to:

[0144] Obtain new document data at preset time intervals;

[0145] Based on the multi-level division rule, the newly added document data is divided and processed to obtain a newly added document division result;

[0146] Based on the newly added document classification result, data updating processing is performed on each of the sub-document title data pools to obtain an updated sub-document title data pool.

[0147] The embodiment of the present invention provides a device for optimizing the citation link speed. Compared with the prior art, the present invention can extract features of different types of data according to different needs by presetting a multi-level division rule with high flexibility. The document title data pool is divided and processed by the multi-level division rule to obtain multiple sub-document title data pools; and the keywords in the citation information to be linked are extracted by the multi-level division rule, and the citation information to be linked is classified and processed based on the keywords to obtain the target citation information category. Then, the association between the citation information category and the sub-document title data pool is obtained, and the target sub-document title data pool corresponding to the target citation information category is determined, so as to optimize the retrieval speed based on the characteristics of the data itself. The present invention reduces the amount of data scanning and improves the single retrieval speed, and finally accelerates the citation link speed, and shares the growing data pressure to each sub-data pool, which greatly alleviates the growing data pressure.

[0148] According to one embodiment of the present invention, a storage medium is provided, wherein the storage medium stores at least one executable instruction, and the computer executable instruction can execute the method for optimizing the citation link speed in any of the above method embodiments.

[0149] Fig. 9 A schematic diagram of the structure of a computer device provided according to an embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computer device.

[0150] like Fig. 9As shown, the computer device may include: a processor (processor) 602 , a communication interface (Communications Interface) 604 , a memory (memory) 606 , and a communication bus 608 .

[0151] The processor 602 , the communication interface 604 , and the memory 606 communicate with each other via a communication bus 608 .

[0152] The communication interface 604 is used to communicate with other devices such as clients or other servers.

[0153] The processor 602 is used to execute the program 610, and specifically can execute the relevant steps of the above-mentioned method for optimizing the citation link speed.

[0154] Specifically, the program 610 may include program codes, which include computer operation instructions.

[0155] The processor 602 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the computer device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0156] The memory 606 is used to store the program 610. The memory 606 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0157] The program 610 may be specifically configured to enable the processor 602 to perform the following operations:

[0158] Obtaining a preset multi-level division rule, and dividing the document title data pool based on the multi-level division rule to obtain a plurality of sub-document title data pools;

[0159] Extracting keywords from the citation information to be linked based on the multi-level division rule, and classifying the citation information to be linked based on the keywords to obtain a target citation information category;

[0160] Acquire the association relationship between the citation information category and the sub-document bibliographic data pool, and determine the target sub-document bibliographic data pool corresponding to the target citation information category based on the association relationship;

[0161] Target document information corresponding to the citation information to be linked is retrieved from the target sub-document title data pool, and a document link is established based on the target document information.

[0162] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0163] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for optimizing citation link speed, characterized in that: include: Obtaining a preset multi-level division rule, and dividing the document title data pool based on the multi-level division rule to obtain a plurality of sub-document title data pools; Extracting keywords from the citation information to be linked based on the multi-level division rule, and classifying the citation information to be linked based on the keywords to obtain a target citation information category; Acquire the association relationship between the citation information category and the sub-document bibliographic data pool, and determine the target sub-document bibliographic data pool corresponding to the target citation information category based on the association relationship; Target document information corresponding to the citation information to be linked is retrieved from the target sub-document title data pool, and a document link is established based on the target document information.

2. The method according to claim 1, characterized in that The document title data pool is divided based on the multi-level division rule to obtain multiple sub-document title data pools, including: Acquire the hierarchical features corresponding to each level in the multi-level division rule; and combine the hierarchical features to obtain document feature information corresponding to each sub-document title data pool; The document information in the document bibliographic data pool is screened and processed in sequence according to the hierarchical order and based on the hierarchical features to obtain the sub-document bibliographic data pool corresponding to each of the document characteristic information.

3. The method according to claim 2, characterized in that The step of screening the document information in the document bibliographic data pool in sequence based on the hierarchical features comprises: Acquire first-level features, and screen the document information in the document bibliographic data pool based on the first-level features to obtain first-level screening results; Acquire second-level features, and perform screening processing on the document information in the first-level screening results based on the second-level features to obtain second-level screening results; A third-level feature is obtained, and based on the third-level feature, the document information in the second-level screening result is screened to obtain a third-level screening result.

4. The method according to claim 1, characterized in that: The step of extracting keywords from the citation information to be linked based on the multi-level division rule includes: Identifying the separation points in the citation information to be linked, and performing block processing on the citation information to be linked based on the separation points to obtain a block result of the citation information to be linked; Obtaining hierarchical features corresponding to each level in the multi-level division rule; The keyword is extracted from the to-be-linked citation block result based on the hierarchical feature.

5. The method according to claim 1, characterized in that Before classifying the citation information to be linked based on the keyword to obtain the target citation information category, the method further includes: A plurality of citation information categories equal in number to the sub-document title data pools are set; the citation information categories have the same document characteristic information as the sub-document title data pools; Acquire a citation information sample data set, and perform block processing on the citation information samples in the citation information sample data set to obtain a block result; Keyword information is extracted from the segmentation results, and a citation classification model is trained based on the keyword information to obtain a citation classification model for identifying the category of the citation information.

6. The method according to claim 1, characterized in that Before obtaining the association relationship between the citation information category and the sub-document bibliographic data pool, the method further includes: Respectively obtain document feature information of the sub-document title data pool; and obtain citation information categories having the same document feature information; The document feature information is associated with the document feature category to obtain an association relationship between the citation information category and the sub-document bibliographic data pool.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: Obtain new document data at preset time intervals; Based on the multi-level division rule, the newly added document data is divided and processed to obtain a newly added document division result; Based on the newly added document classification result, data updating processing is performed on each of the sub-document title data pools to obtain an updated sub-document title data pool.

8. A device for optimizing citation link speed, characterized in that: include: A data pool partitioning module is used to obtain a preset multi-level partitioning rule, and partition the document title data pool based on the multi-level partitioning rule to obtain multiple sub-document title data pools; A citation classification module, used to extract keywords from the citation information to be linked based on the multi-level division rule, and classify the citation information to be linked based on the keywords to obtain a target citation information category; A target determination module, used to obtain the association relationship between the citation information category and the sub-document bibliographic data pool, and determine the target sub-document bibliographic data pool corresponding to the target citation information category based on the association relationship; The retrieval module is used to retrieve the target document information corresponding to the citation information to be linked from the target sub-document title data pool, and establish a document link based on the target document information.

9. A storage medium storing at least one executable instruction, wherein the executable instruction executes an operation corresponding to the method for optimizing citation link speed according to any one of claims 1 to 7.

10. A computer device, comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the method for optimizing the citation link speed according to any one of claims 1-7.

Images