Article search method and device, computer equipment and storage medium

Abstract

The present invention provides an article search method and device, a computer device and a storage medium, wherein the article search method comprises: receiving a search request input by a user through a terminal, wherein the search request carries a search term; determining multiple articles corresponding to the search term based on a database, wherein the database includes a correspondence between articles, index content and category tags, wherein the index content is obtained by performing word segmentation and parsing on the articles, and the category tags are used to indicate the category of the index content; determining the arrangement order of multiple articles according to a predetermined rule; and outputting feedback information to the terminal, wherein the feedback information includes multiple articles and the arrangement order. The problem that the search method in the prior art cannot quickly search for the results desired by the user is solved.

Description

Technical Field

[0001] The present invention relates to the field of Internet technology, and in particular to an article search method and device, computer equipment and storage medium. Background Art

[0002] Searching means that the user inputs a search word or keyword through the terminal, and after receiving the search word sent by the terminal, the server searches for relevant content on the Internet based on the search word and feeds back the search results to the terminal. In the related art, taking books as an example, the server can usually only recognize search words such as book title and author. In this case, once the user forgets the book title and author, it will be difficult to quickly search for the results they want, wasting time and energy. Summary of the invention

[0003] In view of this, an embodiment of the present invention provides an article search method and apparatus, a computer device, and a storage medium to solve the problem that the search method in the prior art cannot quickly search for the results desired by the user.

[0004] The first aspect of the present invention provides an article search method, comprising: receiving a search request input by a user through a terminal, the search request carrying a search term; determining multiple articles corresponding to the search term based on a database, the database including the correspondence between articles, index content and category labels, the index content is obtained by performing word segmentation and parsing on the articles, and the category label is used to indicate the category of the index content; determining an arrangement order of multiple articles according to a predetermined rule; and outputting feedback information to the terminal, the feedback information including multiple articles and the arrangement order.

[0005] In one embodiment, before receiving the search request input by the user through the terminal, the method further includes: obtaining an article; performing word segmentation and parsing on the article to identify words or word sequences in the article; identifying predetermined words in the words or word sequences based on a preset recognition word list, and configuring a category label for the predetermined words. The recognition word list includes a correspondence between the predetermined words and the category label, and the index content includes the predetermined words.

[0006] In one embodiment, the index content includes entity words extracted from the article; and the category tags include entity category tags.

[0007] In one embodiment, before determining the arrangement order of the plurality of articles according to a predetermined rule, the method further includes: determining the search intent score of each of the plurality of articles based on the relevance of at least one of the index content and the category label to the search term, wherein the search intent score is used to indicate the degree of match between the article and the search term. Determining the arrangement order of the plurality of articles according to the predetermined rule includes: determining the arrangement order of the plurality of articles in descending order of the search intent score.

[0008] In one embodiment, before determining the arrangement order of the plurality of articles according to the predetermined rule, the method further includes: determining the search intention score of each of the plurality of articles based on a database, the search intention score being used to indicate the degree of match between the article and the search term; and adjusting the search intention score of each of the plurality of articles based on the user's preference information for the article. Outputting feedback information includes: determining the arrangement order of the plurality of articles according to the predetermined rule includes: determining the arrangement order of the plurality of articles according to the adjusted search intention score in descending order.

[0009] In one embodiment, adjusting the search intent score of each of the multiple articles based on the user's preference information for the articles includes: extracting features of at least one dimension for each of the multiple articles based on a pre-trained matching model; calculating the matching degree between the extracted features and the user's preference information for the articles to obtain the matching degree between each of the multiple articles and the user's preference information for the articles; and adjusting the search intent score based on the matching degree.

[0010] The second aspect of the present invention provides an article search device, including: a receiving module, used to receive a search request input by a user through a terminal, the search request carrying a search term; a first determination module, used to determine multiple articles corresponding to the search term based on a database, the database including the correspondence between articles, index content and category labels, the index content is obtained by performing word segmentation and parsing on the articles, and the category label is used to indicate the category of the index content; a sorting module, used to determine the arrangement order of the multiple articles according to a predetermined rule; a feedback module, used to output feedback information to the terminal, the feedback information including multiple articles and the arrangement order.

[0011] In one embodiment, the article search device also includes: an acquisition module for acquiring articles; a parsing module for performing word segmentation and parsing on the article to identify words or word sequences in the article; a configuration module for identifying predetermined words in the words or word sequences based on a preset recognition word list, and configuring category labels for the predetermined words; the recognition word list includes the correspondence between the predetermined words and the category labels, and the index content includes the predetermined words.

[0012] In one embodiment, the index content includes entity words extracted from the article; and the category tags include entity category tags.

[0013] In one embodiment, the article search device also includes: a second determination module, used to determine the search intent score of each of the multiple articles based on the relevance of at least one of the index content and category tags to the search term, the search intent score being used to indicate the degree of match between the article and the search term; and a sorting module specifically used to determine the arrangement order of the multiple articles in descending order of the search intent score.

[0014] In one embodiment, the article search device also includes: a third determination module, used to determine the search intent score of each of the multiple articles based on a database, the search intent score is used to indicate the degree of match between the article and the search term; an adjustment module, used to adjust the search intent score of each of the multiple articles based on the user's preference information for the articles; and a sorting module is specifically used to determine the arrangement order of the multiple articles in descending order according to the adjusted search intent scores.

[0015] In one embodiment, the adjustment module includes: a feature extraction unit, which is used to extract features of at least one dimension for multiple articles based on a pre-trained matching model; a calculation unit, which is used to calculate the matching degree between the extracted features and the user's preference information for the articles, and obtain the matching degree between each of the multiple articles and the user's preference information for the articles; and an adjustment unit, which is used to adjust the search intent score based on the matching degree.

[0016] A third aspect of the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executed by the processor, wherein the processor implements the steps of the article search method provided in any of the above embodiments when executing the computer program.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the article search method provided in any of the above embodiments are implemented.

[0018] According to the article search method provided by the present invention, a database is established based on the index content obtained by word segmentation and parsing of the article, and the index content is classified according to a plurality of preset category tags to form a database including the correspondence between articles, index content and category tags. The plurality of category tags correspond to a plurality of different dimensions, for example, the plurality of category tags include character names, schools, occupations, nicknames, secrets, moves, weapons, catchphrases, story years, geographical locations, etc., so that the search terms are no longer limited to the two dimensions of book title and author, and multi-dimensional search is realized. In particular, in the scenario where the user forgets the title and author of the book and only remembers key information such as the character names, schools, catchphrases, geographical locations, and story years in the book, it can also ensure that the user can quickly search for the articles they want to find. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of an implementation scenario of an article search method and device provided by an embodiment of the present invention.

[0020] Figure 2 It is a flowchart of an article search method provided by an embodiment of the present invention.

[0021] Figure 3A schematic diagram of the execution process of an article search method provided by an embodiment of the present invention.

[0022] Figure 4 A schematic diagram of the storage structure of articles in a database provided by one embodiment of the present invention.

[0023] Figure 5 A flowchart of an article search method provided by another embodiment of the present invention.

[0024] Figure 6 A flowchart of an article search method provided by another embodiment of the present invention.

[0025] Figure 7 A structural block diagram of user preference information for articles provided by one embodiment of the present invention.

[0026] Figure 8 A structural block diagram of a search device provided by an embodiment of the present invention.

[0027] Fig. 9 This is a structural block diagram of a search device provided in another embodiment of the present invention.

[0028] Fig.10 It is a structural block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. SUMMARY OF THE INVENTION

[0031] In the process of searching for books, the search is usually based on the title and author. It can be seen that the current search method has a narrow scope of application for search terms (or keywords). Once the user cannot provide the search terms that meet the requirements, it will be difficult to find the desired search results. For example, we have all encountered books that we have browsed or read but cannot remember the title and author. We only remember the memorable character names, martial arts or catchphrases in the book. In this case, it is impossible to quickly find the desired search results based on the current search software.

[0032] In view of this, an embodiment of the present invention provides an article search method and device, a computer device, and a storage medium for searching articles. The article mentioned here refers to content presented to users in text form. For example, books, journals, news, papers, etc. According to the article search method and device, computer device, and storage medium provided by an embodiment of the present invention, the server performs word segmentation analysis on the acquired article to obtain multi-dimensional information related to the article, such as character name, school, occupation, nickname, secret book, moves, weapons, catchphrases, story era, geographical location, etc. Then, a database is established based on the multi-dimensional information obtained after word segmentation analysis. Subsequently, when searching for articles, the setting of search terms will not be limited to the title or author of the article, thereby expanding the scope of application of the search terms, and thus ensuring that the search results desired by the user can be found quickly and accurately.

[0033] Exemplary application scenarios

[0034] Figure 1 FIG. 1 is a schematic diagram of an implementation scenario of an article search method and device provided by an embodiment of the present invention. Figure 1 As shown, the implementation scenario 100 includes a terminal 101 , a communication network 102 and a server 103 .

[0035] The terminal 101 may be a mobile phone, a tablet computer, an e-book reader, a laptop computer, a desktop computer, etc. In one embodiment, an application having a search processing function is installed in the terminal 101, and the application has a function of displaying search results. The application mentioned here includes but is not limited to an e-book reading website, an e-book reading APP, etc.

[0036] The terminal 101 and the server 103 are connected via a communication network 102, so that a user can use the terminal 101 to interact with the server 103 via the communication network 102 to receive or send messages. The communication network 102 is a wired network or a wireless network.

[0037] The server 103 is also called a search server or a search engine, and includes a server with a search processing function. The server 103 is a server or a cluster of multiple servers, or a virtualization platform, or a cloud computing service center.

[0038] Taking an e-book reading APP as an example, under normal circumstances, when the server 103 receives a search request carrying a search term sent by the terminal 101, it executes the article search method provided in the following embodiments and sends feedback results to the terminal 101 for display through the user interface of the terminal 101.

[0039] It should be noted that Figure 1The implementation scenario shown includes at least one terminal 101 and at least one server 103. Figure 1 Schematically shows multiple terminals 101 and one server 103, which is not limited in this embodiment.

[0040] The communication network 102 uses standard communication technology and / or protocols. The communication network 102 can be at least one or any combination of the Internet, a local area network (LAW), a metropolitan area network (MAN), a wide area network, a private network, or a virtual private network. In some embodiments, technologies and / or formats including Hyper Text Mark-up Language (HTML), Extensible Markup Language (XML), etc. are used to represent data exchanged through the network. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above data communication technologies.

[0041] Exemplary Methods

[0042] Figure 2 It is a flowchart of an article search method provided by an embodiment of the present invention. Figure 3 A schematic diagram of the execution process of an article search method provided by an exemplary embodiment of the present invention. This embodiment can be applied to Figure 1 In the server 103 shown, the Figure 2 and Figure 3 It can be seen that the article search method includes:

[0043] Step S120: receiving a search request input by a user through a terminal, wherein the search request carries a search term.

[0044] In the case where the user has logged into an application with a search processing function, when the user inputs a search word in the application through the terminal 101 , the terminal 101 sends a search request including the search word to the server 103 .

[0045] In one example, the search request includes a novel search request for requesting to search for a target novel in a novel database.

[0046] Step S140, determining multiple articles corresponding to the search term based on a database, the database including the correspondence between articles, index content and category labels, the index content is obtained by performing word segmentation and parsing on the articles, and the category label is used to indicate the category of the index content.

[0047] In one embodiment, the index content includes entity words extracted from the article; accordingly, the category label includes entity category labels. For example, the category label corresponding to the index content "doctor" is "occupation"; the category label corresponding to the index content "Songshan sect" is "sect". The category label corresponding to the same search term can be one or more. For example, the category label corresponding to the search term "sea" can be a geographical name or a character name.

[0048] In one embodiment, the process of establishing a database includes: performing word segmentation analysis on the article to identify words or word sequences in the article; then, identifying predetermined words in the words or word sequences based on a preset recognition word list, and configuring category tags for the predetermined words. Specifically, Figure 4 A schematic diagram of the storage structure of articles in a database provided by an embodiment of the present invention. Figure 4 , the process of word segmentation and parsing of an article includes: when the server obtains an article, the article is segmented and parsed to reassemble the continuous subsequences in the article into words or word sequences according to certain specifications. Then, in combination with the preset entity recognition vocabulary, the entity words in the article body are identified, and corresponding entity category tags are configured for these entity words. Among them, the preset entity recognition vocabulary includes the correspondence between entity words and entity category tags, and each entity category tag corresponds to a predetermined matching rule. When the word or word sequence parsed from the article body meets the corresponding matching rule, the server will determine the word or word sequence as an entity word and configure the corresponding entity category tag for the entity word. In one embodiment, the entity category tag includes at least one of: character name, school, occupation, nickname, secret book, moves, weapons, catchphrases, story era, and geographical location.

[0049] Compared to characters, words have clear semantics, so they can reflect the user's search intent and can be used as search terms. However, the number of entity words identified from the article body is huge, that is, the index content in the database is huge. Therefore, in order to find index content related to the search term from the database, the traversal search method is inefficient. Therefore, by classifying the index content and formulating corresponding category labels for each type, when determining multiple articles corresponding to the search term based on the database later, you can first determine the category label to which the search term belongs, and then search for index content related to the search term under the category label, which can greatly reduce the amount of calculation.

[0050] When searching for index content related to the search term under a category label, a preset matching algorithm can be used to determine the index content related to the search term under the category label. The preset matching algorithm is, for example, Linear Discriminant Analysis (LDA), Deep Structured Semantic Models (DSSM), Convolutional Deep Structured Semantic Models (CDSSM), Deep Relevance Matching Model (DRMM), etc. After the calculation process of the preset matching algorithm, the articles corresponding to the index content with a matching degree greater than the preset threshold are determined as search results.

[0051] Step S160, determining the arrangement order of the plurality of articles according to a predetermined rule.

[0052] Step S180, output feedback information to the terminal, the feedback information includes multiple articles and the arrangement order. Subsequently, the terminal can display the prompt information of each of the multiple articles in the arrangement order according to the feedback information. The prompt information can include the matched words or word sequences, article category labels, and can also include article titles, authors, etc. Among them, the information such as article category labels, article titles, and authors can be directly obtained when obtaining the articles.

[0053] According to the article search method provided by this embodiment, a database is established based on the index content obtained by word segmentation and parsing of the article, and the index content is classified according to a plurality of preset category tags to form a database including the correspondence between articles, index content and category tags. The plurality of category tags correspond to a plurality of different dimensions respectively. For example, the plurality of category tags include character names, schools, occupations, nicknames, secrets, moves, weapons, catchphrases, story years, and geographical locations, so that the search terms are no longer limited to the two dimensions of book title and author, and multi-dimensional search is realized. In particular, in the scenario where the user forgets the title and author of the book and only remembers key information such as the character names, schools, catchphrases, geographical locations, and story years in the book, it can also ensure that the user can quickly search for the articles they want to find.

[0054] Figure 5 FIG. 1 is a flow chart of an article search method provided by another embodiment of the present invention. Figure 5 The article search method 200 shown and Figure 2 The difference of the article search method 100 is that, in this embodiment, before executing step S160, it also includes:

[0055] Step S250, based on the relevance of at least one of the index content and the category label to the search term, determine the search intent score of each of the multiple articles, where the search intent score is used to indicate the degree of match between the article and the search term.

[0056] In one embodiment, based on the text relevance between the index content and the search term, the search intent score of each of the multiple articles is determined. For example, for the search term "wu xie", the index content matched according to the preset matching algorithm includes "Wu Xie" and "Wu Xie", then the search intent score corresponding to "Wu Xie" is equal to the search intent score corresponding to "Wu Xie". It should be understood that when determining the matching degree between the index content and the search term, it is not limited to the method of determining the text relevance between the index content and the search term. In specific applications, other factors, such as the average number of clicks by users, etc., can also be combined. Still taking the search term "wuxie" as an example, when the index content matched by the search term "wu xie" according to the preset matching algorithm is "Wu Xie" and "Wu Xie", and the text relevance between "Wu Xie" and "Wu Xie" and the search term "wu xie" is equal, the average number of clicks by users of "Wu Xie" and "Wu Xie" can be further determined. When the average number of clicks by users of "Wu Xie" is greater than that of "Wu Xie", it is determined that the search intent score corresponding to "Wu Xie" is higher than the search intent score corresponding to "Wu Xie".

[0057] In one embodiment, the search intent scores of the multiple articles are determined based on the relevance between the category tags and the search terms. For example, the category tags corresponding to the search term "wu xie" include "book title" and "character name", wherein 50 articles corresponding to the index content related to the search term are matched under the category tag "book title", and 100 articles corresponding to the index content related to the search term are matched under the category tag "character name", then the relevance between the category tag "book title" and the search term is higher than the relevance between the category tag "character name" and the search term.

[0058] In one embodiment, the search intent scores of the multiple articles are determined based on the relevance of at least one of the index content and the category label to the search term. For example, the server scores the relevance of the index content and the search term, and the relevance of the category label and the search term, respectively, and then performs a weighted sum of the two scores to obtain a comprehensive score as the search intent score.

[0059] In this case, step S160 is specifically performed as follows:

[0060] Step S260, determining the order of arranging the multiple articles in descending order of the search intent score, that is, arranging the multiple articles in descending order of the search intent score, with the articles with high search intent scores at the front and the articles with low search intent scores at the back.

[0061] The higher the search intent score, the more relevant the article is to the search term; the lower the search intent score, the less relevant the article is to the search term. The terminal displays the prompt information of multiple articles in descending order of search intent scores, ensuring that the articles with higher relevance to the search term are displayed in the front position.

[0062] In one embodiment, when there are at least two articles with equal search intent scores among the plurality of articles, the at least two articles may be displayed in a random order.

[0063] Figure 6 FIG. 1 is a flow chart of an article search method provided by another embodiment of the present invention. Figure 6 The article search method 300 shown and Figure 2 The difference of the article search method 100 shown is that, in this embodiment, referring to Figure 6 and Figure 3 , before executing step S160, further comprising:

[0064] Step S351, determining the search intent score of each of the plurality of articles based on the database, where the search intent score is used to indicate the degree of match between the article and the search term.

[0065] The specific execution process of determining the search intent scores of the multiple articles can refer to the execution process of step S250, which will not be repeated here.

[0066] Step S352: adjusting the search intent scores of the multiple articles based on the user's preference information for the articles.

[0067] The user's preference information for articles can be obtained in the following manner: In one example, the server first determines the terminal identifier corresponding to the search request, and then determines the user account identifier based on the pre-stored correspondence between the terminal identifier and the user account identifier, and then determines the user's preference information for articles corresponding to the user account identifier.

[0068] The search request itself carries a terminal identifier, which is used to uniquely identify the terminal. For example, the terminal identifier is the International Mobile Equipment Identity (IMEI). Usually, the terminal identifier corresponds to at least one user account identifier. When at least one user logs in to the application through the same terminal, the server automatically establishes a correspondence between the terminal identifier and the user account identifier. The user account identifier is used to uniquely identify the user login account. For example, the user account identifier is the user login account. The user account identifier corresponds one-to-one to the user's preference information for the article.

[0069] In this way, by first determining the terminal identifier corresponding to the search request and then determining the user account identifier corresponding to the terminal identifier, compared to directly determining the user account identifier corresponding to the search request, the terminal identifier is used to filter out a large number of irrelevant user account identifiers, thereby greatly improving the retrieval efficiency.

[0070] In another example, the search request also carries a user account identifier. In this case, the server directly extracts the user account identifier from the received search request, and then obtains the user's preference information for the article based on the user account identifier.

[0071] In one embodiment, the server can use a pre-trained matching model to extract features of at least one dimension for multiple articles determined based on the search terms, and then calculate the matching degree between the extracted features and the user's preference information for the articles to obtain the matching degree between each of the multiple articles and the user's preference information for the articles; and then use the matching degree to adjust the search intent score.

[0072] The pre-trained matching models include but are not limited to: at least one of a deep neural network model (DNN), a recurrent neural network model (RNN), an embedding model, a gradient boosting decision tree (GBDT) model, and a logistic regression (LR) model.

[0073] Take a dimension, such as article category, for example. Figure 2 , the database also includes article category tags, which are used to identify the category to which the article belongs. Taking novels as an example, the article category tags set for novels include fantasy, suspense, romance, history, war, time travel, etc. Suppose user A prefers romance novels, and user B prefers suspense novels. When user A and user B search with the search term "wu xie" respectively, the multiple articles determined based on the database and the search intent scores of the multiple articles are the same. However, after the user's preference information for the article is adjusted, the first article displayed on the terminal where user A is located is "Wu Xie" by the author Yinshan, and the first article displayed on the terminal where user B is located is "Grave Robbers' Chronicles" by Nanpai Sanshu (the protagonist's name is "Wu Xie").

[0074] In one embodiment, see Figure 4 , article category labels can be classified in a hierarchical manner. The specific classification method can be reasonably set according to actual needs, for example, article category labels can be refined into three levels.

[0075] In one embodiment, the process of adjusting the search intent score using the degree of match between each of multiple articles and the user's preference information for the articles includes: determining the degree of match as a match score; taking a weighted sum of the match score and the search intent score, and obtaining a comprehensive score as the adjusted search intent score.

[0076] In one embodiment, according to step S351, a first article list is obtained, and the articles in the first article list are arranged in descending order according to the search intent score; according to step S352, a second article list is obtained, and the articles in the second article list are arranged in descending order according to the adjusted search intent score.

[0077] In this case, step S160 is specifically performed as follows:

[0078] Step S360, determining the arrangement order of the plurality of articles according to the adjusted search intent scores from high to low.

[0079] In this case, for user A who prefers romance novels, even if the search term entered by user A is "Wu Xie", the server will rank "Wu Xie" by author Yin Shan before "Grave Robbers' Chronicles" (the protagonist's name is "Wu Xie"). This can be applied to the following scenario: user A originally wanted to search for "Wu Xie" by author Yin Shan, but user A mistakenly entered "Wu Xie" as "Wu Xie", thereby ensuring that the search results that the user wants are displayed at the top as much as possible.

[0080] According to the article search method provided in this embodiment, the search intent score is adjusted by utilizing the user's preference information for articles, so that for the same search term, the priority of displaying articles to different users is different, thereby achieving personalized prompts.

[0081] Figure 7 A structural block diagram of user preference information for articles provided by an embodiment of the present invention. Figure 7 As shown, the user's preference information 60 for articles includes multiple user tags, multiple reference dimensions corresponding to each of the multiple user tags, and multiple dimensional features corresponding to each of the multiple reference dimensions. Dimensional features refer to specific values ​​or categories corresponding to the reference dimensions.

[0082] In this embodiment, user tags include behavior tags, model tags, and advanced tags. The reference dimensions corresponding to the behavior tags include the number of times the APP is used, the number of active days, the categories of books browsed, the categories of books read, the reading time, and the reading progress. The reference dimensions corresponding to the model tags include demographic attributes, geographic location, interests and hobbies, product preferences, browsing content other than novels, and the browsing time of content other than novels. The reference dimensions corresponding to the advanced tags include population attributes, whether they have been lost, and conversion potential.

[0083] When adjusting the search intent scores of multiple articles based on the user's preference information for the articles, the server extracts the dimensional features of each article under at least one reference dimension through a pre-trained matching model. In one example, the at least one reference dimension includes at least one of the article category, story year, crowd age, and population attribute, and this information can be directly obtained when the article is obtained; then the extracted dimensional features and the corresponding dimensional features in the user's preference information for the articles are matched, thereby obtaining the matching degree between each of the multiple articles and the user's preference information for the articles.

[0084] For example, the matching model calculates the matching degree from three reference dimensions: book classification, demographic attributes, and crowd attributes. Then the database should store the dimensional features of the articles in these three reference dimensions. For example, the dimensional features of article a in these four reference dimensions are: romance, female, and age group 18 to 30. The dimensional features of article b in these four reference dimensions are: fantasy, male, and age group 18 to 30. The dimensional features of user A's user preference information for articles in these four reference dimensions are: romance, female, and age group 18 to 30. Then the matching degree between "article a" and "user preference information for articles" output by the matching model is greater than the matching degree between "article b" and "user preference information for articles".

[0085] It should be understood that the reference dimension selected by the matching model can be reasonably set according to actual needs. The specific structure of the user's preference information for articles can also be reasonably set according to actual needs.

[0086] Exemplary Devices

[0087] The invention also provides an article search device. Figure 8 FIG. 1 is a block diagram of a search device provided for a first exemplary embodiment of the present invention. Figure 8 As shown, the search device 70 includes a receiving module 71, a first determination module 72, a sorting module 73 and a feedback module 74. Among them, the receiving module 71 is used to receive a search request input by a user through a terminal, and the search request carries a search term. The first determination module 72 is used to determine multiple articles corresponding to the search term based on a database, and the database includes the correspondence between articles, index content and category labels. The index content is obtained by performing word segmentation and parsing on the article, and the category label is used to indicate the category of the index content. The sorting module 73 is used to determine the arrangement order of multiple articles according to a predetermined rule. The feedback module 74 is used to output feedback information to the terminal, and the feedback information includes multiple articles and the arrangement order.

[0088] In one embodiment, the index content includes entity words extracted from the article; accordingly, the category tags include entity category tags.

[0089] In one embodiment, the search device 70 further includes: an acquisition module, a parsing module and a configuration module. The acquisition module is used to acquire an article. The parsing module is used to perform word segmentation and parsing on the article to identify words or word sequences in the article. The configuration module is used to identify predetermined words in the words or word sequences based on a preset recognition word list and configure a category label for the predetermined words.

[0090] In one embodiment, the entity category tag includes at least one of: character name, school, occupation, nickname, secret book, moves, weapon, catchphrase, story era, and geographical location.

[0091] In one embodiment, the article search device 70 further includes: a second determination module, which is used to determine the search intention score of each of the plurality of articles based on the database, and the search intention score is used to indicate the degree of match between the article and the search term. For example, the second determination module is specifically used to determine the search intention score of each of the plurality of articles based on the degree of relevance of at least one of the index content and the category label to the search term, and the search intention score is used to indicate the degree of match between the article and the search term.

[0092] In this case, the sorting module 73 is specifically used to determine the arrangement order of multiple articles according to the order of search intent scores from high to low.

[0093] According to the article search device provided by this embodiment, a database is established based on the index content obtained by word segmentation and parsing of the article, and the index content is classified according to a plurality of preset category tags to form a database including the correspondence between articles, index content and category tags. The plurality of category tags correspond to a plurality of different dimensions respectively. For example, the plurality of category tags include character names, schools, occupations, nicknames, secrets, moves, weapons, catchphrases, story years, and geographical locations, so that the search terms are no longer limited to the two dimensions of book title and author, and multi-dimensional search is realized. In particular, in the scenario where the user forgets the title and author of the book and only remembers key information such as the character names, schools, catchphrases, geographical locations, and story years in the book, it can also ensure that the user can quickly search for the articles they want to find.

[0094] Fig. 9 FIG. 1 is a structural block diagram of a search device provided by another embodiment of the present invention. Fig. 9 The search device 80 shown is Figure 8 The search device 70 shown further includes: a third determination module 81 and an adjustment module 82. The third determination module 81 is used to determine the search intention score of each of the multiple articles based on the database, and the search intention score is used to indicate the degree of match between the article and the search term. The adjustment module 82 is used to adjust the search intention score of each of the multiple articles based on the user's preference information for the article. In this case, the sorting module 73 is used to determine the arrangement order of the multiple articles according to the adjusted search intention score from high to low.

[0095] In one example, the adjustment module 82 specifically includes a feature extraction unit, a calculation unit, and an adjustment unit. The feature extraction unit is used to extract features of at least one dimension for each of the multiple articles determined based on the search term based on a pre-trained matching model. The calculation unit is used to calculate the matching degree between the extracted features and the user's preference information for the article, and obtain the matching degree between each of the multiple articles and the user's preference information for the article. The adjustment unit is used to adjust the search intent score based on the matching degree.

[0096] In one example, the adjustment unit is specifically used to determine the matching degree as a matching score; perform weighted summation on the matching score and the search intent score, and obtain a comprehensive score as the adjusted search intent score.

[0097] The article search device provided in this embodiment belongs to the same inventive concept as the article search method provided in the embodiment of the present invention, can execute the article search method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the article search method. For technical details not fully described in this embodiment, please refer to the article search method provided in the embodiment of the present invention, and will not be repeated here.

[0098] Exemplary Electronic Devices

[0099] Below, reference Fig.10 The electronic device provided by the embodiment of the present invention is described as follows. The electronic device may be Figure 1 Server 103 in. Fig.10 FIG. 1 is a block diagram of an electronic device provided by an embodiment of the present invention. Fig.10 As shown, the electronic device 10 includes one or more processors 11 and a memory 12 .

[0100] The processor 11 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

[0101] The memory 12 may include one or more computer program products, and the computer program product may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may run the program instructions to implement the article search method of each embodiment of the present invention described above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.

[0102] In one example, the electronic device 10 may further include: an input device 13 and an output device 14 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0103] For example, the input device 13 may be a microphone or a microphone array for capturing input signals from a sound source. When the electronic device is a stand-alone device, the input device 13 may be a connector of the communication network 102 for receiving collected input signals from the terminal 101. In addition, the input device 13 may also include, for example, a keyboard, a mouse, and the like.

[0104] The output device 14 can output various information to the outside, including the determined distance information, direction information, etc. The output device 14 can include, for example, a display, a speaker, a printer, and the communication network 102 and the remote output devices connected thereto.

[0105] Of course, to simplify, Fig.10 Only some of the components related to the present invention in the electronic device 10 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application conditions, the electronic device 10 may also include any other appropriate components.

[0106] Exemplary computer program products and computer-readable storage media

[0107] In addition to the above-mentioned methods and devices, an embodiment of the present invention may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the article search method according to various embodiments of the present invention described in the above "Exemplary Method" section of this specification.

[0108] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the embodiments of the present invention, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0109] In addition, an embodiment of the present invention may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by the processor, the processor 11 executes the steps of the article search method according to various embodiments of the present invention described in the above "Exemplary Method" section of this specification.

[0110] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0111] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details disclosed above are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0112] The block diagrams of the devices, apparatuses, equipment, and systems involved in the present invention are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The word "such as" used here refers to the phrase "such as but not limited to", and can be used interchangeably with it.

[0113] It should also be noted that in the device, apparatus and method of the present invention, each component or each step can be decomposed and / or reassembled, and such decomposition and / or reassembly should be regarded as an equivalent solution of the present invention.

[0114] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0115] It should be understood that the qualifiers "first", "second", "third", "fourth", "fifth" and "sixth" used in the description of the embodiments of the present invention are only used to more clearly explain the technical solutions and cannot be used to limit the scope of protection of the present invention.

[0116] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present invention to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. An article search method, characterized in that: include: Receiving a search request input by a user through a terminal, wherein the search request carries a search term; Determine the category label to which the search term belongs, search for index content related to the search term in a database based on the category label, and determine multiple articles corresponding to the search term in the database based on the searched index content; The database is obtained by classifying multiple index contents based on multiple preset category tags, the multiple index contents are obtained by word segmentation and parsing of articles, the database includes the correspondence between the articles, the index contents and the category tags, and the category tags are used to indicate the category of the index contents; Determining a search intent score for each of the plurality of articles based on a relevance of at least one of the index content and the category label to the search term, the search intent score being used to indicate a matching degree between the article and the search term; Determining an arrangement order of the plurality of articles according to the order of the search intent scores from high to low; Outputting feedback information to the terminal, the feedback information including the plurality of articles and the arrangement order.

2. The article search method according to claim 1, characterized in that: Before receiving the search request input by the user through the terminal, the method further includes: obtaining said article; Performing word segmentation analysis on the article to identify words or word sequences in the article; Recognize a predetermined word in the word or word sequence based on a preset recognition word list, and configure the category label for the predetermined word; The recognition word list includes the correspondence between the predetermined words and the category labels, and the index content includes the predetermined words.

3. The article search method according to claim 1, characterized in that: The index content includes entity words extracted from the article; the category labels include entity category labels.

4. The article search method according to any one of claims 1 to 3, characterized in that: Before determining the arrangement order of the plurality of articles in descending order of the search intent scores, the method further includes: Based on the user's preference information for the articles, adjusting the search intent scores of the plurality of articles; Wherein, determining the arrangement order of the plurality of articles according to the order of the search intent scores from high to low includes: The arrangement order of the plurality of articles is determined according to the adjusted search intent scores from high to low.

5. The article search method according to claim 4, characterized in that: The adjusting the search intention scores of the respective articles based on the user's preference information for the articles includes: Based on the pre-trained matching model, extracting features of at least one dimension from each of the multiple articles; Calculating the matching degree between the extracted features and the user's preference information for the articles, and obtaining the matching degree between each of the plurality of articles and the user's preference information for the articles; The search intent score is adjusted based on the matching degree.

6. An article search device, characterized in that: include: A receiving module, configured to receive a search request input by a user through a terminal, wherein the search request carries a search term; A first determination module is used to determine the category label to which the search term belongs, search for index content related to the search term in a database based on the category label, and determine multiple articles corresponding to the search term in the database based on the found index content; wherein the database is obtained by classifying multiple index contents based on multiple preset category labels, the multiple index contents are obtained by performing word segmentation and parsing on articles, the database includes the correspondence between the articles, the index contents and the category label, and the category label is used to indicate the category of the index content; A second determination module, configured to determine a search intent score for each of the plurality of articles based on a relevance between the index content and at least one of the category tags and the search term, wherein the search intent score is used to indicate a matching degree between the article and the search term; A sorting module, used to determine the order of arrangement of the plurality of articles according to the order of the search intent scores from high to low; A feedback module is used to output feedback information to the terminal, wherein the feedback information includes the plurality of articles and the arrangement order.

7. The article search device according to claim 6, characterized in that: Also includes: An acquisition module, used for acquiring the article; A parsing module, used for performing word segmentation analysis on the article to identify words or word sequences in the article; A configuration module is used to identify predetermined words in the word or word sequence based on a preset recognition word table and configure the category label for the predetermined word; the recognition word table includes the correspondence between the predetermined word and the category label, and the index content includes the predetermined word.

8. The article search device according to claim 6, characterized in that: The index content includes entity words extracted from the article; the category labels include entity category labels.

9. The article search device according to any one of claims 6 to 8, characterized in that: Also includes: An adjustment module, configured to adjust the search intent scores of each of the plurality of articles based on the user's preference information for the articles; The sorting module is specifically used to determine the arrangement order of the multiple articles according to the adjusted search intent scores from high to low.

10. The article search device according to claim 9, characterized in that: The adjustment module comprises: A feature extraction unit, configured to extract features of at least one dimension from each of the plurality of articles based on a pre-trained matching model; A calculation unit, configured to calculate a matching degree between the extracted features and the user's preference information for the article, to obtain a matching degree between each of the plurality of articles and the user's preference information for the article; An adjustment unit is used to adjust the search intent score based on the matching degree.

11. A computer device comprising a memory, a processor, and a computer program stored in the memory and executed by the processor, characterized in that: When the processor executes the computer program, the steps of the article search method according to any one of claims 1 to 5 are implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the article search method according to any one of claims 1 to 5 are implemented.

13. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, the article search method according to any one of claims 1 to 5 is implemented.

Images