WO2020057022A1 - 关联推荐方法、装置、计算机设备和存储介质 - Google Patents

关联推荐方法、装置、计算机设备和存储介质 Download PDF

Info

Publication number
WO2020057022A1
WO2020057022A1 PCT/CN2019/071180 CN2019071180W WO2020057022A1 WO 2020057022 A1 WO2020057022 A1 WO 2020057022A1 CN 2019071180 W CN2019071180 W CN 2019071180W WO 2020057022 A1 WO2020057022 A1 WO 2020057022A1
Authority
WO
WIPO (PCT)
Prior art keywords
search
user
keywords
candidate
keyword
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2019/071180
Other languages
English (en)
French (fr)
Inventor
江琳
蔡健
赵云松
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
OneConnect Smart Technology Co Ltd
Original Assignee
OneConnect Smart Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by OneConnect Smart Technology Co Ltd filed Critical OneConnect Smart Technology Co Ltd
Priority to KR1020217003560A priority Critical patent/KR20210024173A/ko
Priority to SG11202101614VA priority patent/SG11202101614VA/en
Priority to JP2021504196A priority patent/JP7073576B2/ja
Priority to EP19863131.9A priority patent/EP3855324A4/en
Publication of WO2020057022A1 publication Critical patent/WO2020057022A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/906Clustering; Classification

Definitions

  • the present application relates to an association recommendation method, apparatus, computer equipment, and storage medium.
  • the inventors realized that traditional search technologies, when a user enters a query word through a search engine, can automatically recommend candidate words under the search box, and recommend other queries related to the user input query semantics to the user.
  • the traditional search engine of the major search engines is based on the search enthusiasm of individual keywords and personal search history. It cannot be combined with user search intent to push service information that meets their own needs, and cannot be associated with user permissions and associated groups. , Recommendation is poorly related.
  • an association recommendation method is provided.
  • An association recommendation method includes:
  • the search request Receiving a search request sent by a terminal, and obtaining a search behavior log of a user according to the search request; the search request carrying a user identifier, the user identifier corresponding to a user right;
  • a search keyword category According to a preset correspondence relationship between a search keyword category and a related recommendation set, obtain related data corresponding to the search keyword category from the related recommendation set, and send the related data to a terminal.
  • An associated recommendation device includes:
  • a search log obtaining module configured to receive a search request sent by a terminal, and obtain a search behavior log of a user according to the search request; the search request carries a user identifier, and the user identifier corresponds to a user right;
  • a search keyword extraction module configured to extract a corresponding search keyword from the search behavior log according to the user identifier
  • a classification module configured to classify the search keywords according to a preset search keyword category according to the user authority
  • An association data acquisition module configured to obtain association data corresponding to the search keyword category from the association recommendation set according to a preset correspondence relationship between the search keyword category and the association recommendation set, and to obtain the association data Send to terminal.
  • a computer device includes a memory and one or more processors.
  • the memory stores computer-readable instructions.
  • the one or more processors are executed. The following steps:
  • the search request Receiving a search request sent by a terminal, and obtaining a search behavior log of a user according to the search request; the search request carrying a user identifier, the user identifier corresponding to a user right;
  • a search keyword category According to a preset correspondence relationship between a search keyword category and a related recommendation set, obtain related data corresponding to the search keyword category from the related recommendation set, and send the related data to a terminal.
  • One or more non-volatile computer-readable storage media storing computer-readable instructions.
  • the one or more processors execute the following steps:
  • the search request Receiving a search request sent by a terminal, and obtaining a search behavior log of a user according to the search request; the search request carrying a user identifier, the user identifier corresponding to a user right;
  • a search keyword category According to a preset correspondence relationship between a search keyword category and a related recommendation set, obtain related data corresponding to the search keyword category from the related recommendation set, and send the related data to a terminal.
  • FIG. 1 is an application scenario diagram of an association recommendation method according to one or more embodiments.
  • FIG. 2 is a schematic flowchart of an association recommendation method according to one or more embodiments.
  • FIG. 3 is a schematic flowchart of acquiring related data corresponding to a search keyword category according to one or more embodiments.
  • FIG. 4 is a block diagram of an association recommendation device according to one or more embodiments.
  • FIG. 5 is a block diagram of a computer device according to one or more embodiments.
  • the association recommendation method provided in this application can be applied to the application environment shown in FIG. 1.
  • the terminal 102 and the server 104 communicate through a network.
  • the server 104 receives the search request sent by the terminal 102 and obtains a search behavior log of the user according to the search request.
  • the search request carries a user ID, and the user ID corresponds to the user authority.
  • the server 104 extracts the corresponding search keywords from the search behavior log according to the user identification, and classifies the search keywords according to a preset search keyword category according to the user authority.
  • the server 104 obtains the association data corresponding to the search keyword category from the association recommendation set according to the preset correspondence relationship between the search keyword category and the association recommendation set, and sends the association data to the terminal 102.
  • the terminal 102 may be, but is not limited to, various personal computers, notebook computers, smart phones, and tablet computers.
  • the server 104 may be implemented by an independent server or a server cluster composed of multiple servers.
  • an association recommendation method is provided.
  • the method is applied to the server in FIG. 1 as an example, and includes the following steps:
  • the server receives a search request sent by the terminal, and obtains a search behavior log of the user according to the search request.
  • the search request carries a user ID, and the user ID corresponds to the user authority.
  • the search request carries a user identifier, which is used to indicate that the corresponding user performs a search operation on the terminal and enters a corresponding search phrase
  • the server obtains the user's search behavior log from the database according to the search phrase entered by the user.
  • the user ID corresponds to the user authority, and the server may obtain the user authority possessed by the corresponding user according to the user identifier carried in the search request.
  • the search phrase can be a single word or a phrase or sentence composed of multiple words. It can also be Chinese or Pinyin.
  • the search phrase can be "Shenzhen”, "shenzhen”, and "Shanghai people's education level distribution”.
  • the user's search behavior log includes the user's historical search operation and search results corresponding to the historical operation.
  • the corresponding search results include input search results and click search results.
  • the search behavior and corresponding history input search results can be entered based on the user's history and the user's history. Click the search input and the corresponding history to click the search results to generate the corresponding user's search behavior log.
  • the server receives the search request from the terminal, and obtains the user identifier carried by the search request, and obtains the user authority corresponding to the user identifier according to the correspondence between the user identifier and the user authority.
  • the historical search operation and the search result corresponding to the historical search operation corresponding to the user ID are obtained from the database.
  • the historical search operations and Corresponding search results are obtained from the database.
  • the search phrase entered by the user is "How are the academic qualifications of male borrowers in Shanghai distributed?" Shanghai “," Male Borrowers “,” Educational Distribution “,” Shanghai Men's Educational Distribution “, and” Borrower Educational Distribution “, etc., perform multiple historical search operations, and obtain search results corresponding to historical search operations from the database. According to a plurality of historical operations and corresponding search results, a search behavior log of a user corresponding to "What is the education distribution of male borrowers in Shanghai" input by the user is generated.
  • the server extracts a corresponding search keyword from the search behavior log according to the user identifier.
  • the server obtains the corresponding scene according to the user identifier, and according to the search phrase input by the user, obtains the historical search input corresponding to the search phrase in the corresponding scene from the database, and obtains the search corresponding to the historical search input from the search behavior log. result.
  • the server extracts search keywords from the historical search data in the corresponding scene according to the user identification.
  • the user's search behavior log includes the user's historical search operation and search results corresponding to the historical operation.
  • the search keyword is a keyword that is the same as or similar to any one of the search phrases entered by the user in the terminal.
  • the server may obtain a corresponding user behavior search log according to the search phrase entered by the user, and obtain the corresponding user behavior search log according to the user identifier. In the behavior search log for search keywords with the same or similar search phrase.
  • the input search by the user is taken as an example.
  • the search phrase entered by the user is "how is the education distribution of Shenzhen small loan borrowers", and the corresponding historical search operation in the user search behavior log includes “Shenzhen”, “small loan borrowers”, “education distribution”, “Shenzhen small Loan borrowers "and” educational distribution of small loan borrowers ".
  • the search keywords obtained from the corresponding user search behavior log can be: “Shenzhen small loan borrower", “Small loan borrower education distribution”, and "Shenzhen”.
  • the server sorts the search keywords according to a preset search keyword category according to the user authority.
  • the server can obtain a user identity category corresponding to the user authority by acquiring a correspondence between the user authority and the user identifier.
  • the server can obtain a search keyword corresponding to the user identification category by acquiring the correspondence between the user identification category and the search keyword.
  • the server can classify the search keywords corresponding to different user identification categories into different search keyword categories by acquiring the association relationship between the search keyword categories and the search keywords.
  • User rights indicate the rights that the user has on the corresponding terminal, including search, add, delete, and update. Different users can have the same user right. For example, user A has search and update permissions on the terminal, user B has search permissions on the terminal, and user C has update permissions on the terminal.
  • User ID categories represent different users with the same user rights, and the corresponding categories of IDs, that is, user A and user B have search rights, but user B does not have update rights, so user A and user B correspond to different User ID category.
  • search keywords corresponding to different user identification categories may be classified into different search keyword categories. For example, if the obtained search keyword entered by the user A is “Shenzhen University Student Loan Amount Distribution”, according to the search keyword entered by the user A, the corresponding search keyword category can be obtained as “loan”. According to the user identification category corresponding to the user A, the obtained search keyword “Shenzhen University Student Loan Amount Distribution” input by the user A is classified into the search keyword category corresponding to the “loan”.
  • the server obtains the association data corresponding to the search keyword category from the association recommendation set according to the preset correspondence relationship between the search keyword category and the association recommendation set, and sends the association data to the terminal.
  • the server extracts candidate search keywords that match the preset similarity from the database according to the search keywords corresponding to the search keyword category.
  • a search-related candidate vocabulary is generated according to candidate search keywords matching a preset similarity, and a search result corresponding to the candidate search keywords is obtained.
  • a related recommendation set is generated, and the candidate search keywords matching the preset similarity and the search results corresponding to the candidate search keywords are obtained from the related recommendation set.
  • the server calculates the similarity between the search keywords and other search keywords in the database, and obtains candidate search keywords that match the preset similarity, and generates search association candidates according to the candidate search keywords that match the preset similarity Thesaurus.
  • the preset similarity may be set to 0.7, that is, other search keywords corresponding to the similarity of the search keywords of 0.7 may be selected as candidate search keywords, according to a preset similarity greater than or equal to Candidate search keywords with a degree of 0.7 to generate search related candidate thesaurus.
  • the server may generate a related recommendation set based on the obtained search results corresponding to the candidate search keywords, and the search related candidate thesaurus generated based on the candidate search keywords. Further, the server obtains the association data corresponding to the search keyword category from the association recommendation set according to the preset correspondence relationship between the search keyword category and the association recommendation set.
  • the association data is composed of candidate search keywords that match a preset similarity and search results corresponding to the candidate search keywords.
  • the server obtains a search behavior log of a user from a database according to a search request sent by a terminal, and the search request carries a user identifier, and the user identifier corresponds to a user right.
  • the corresponding search keywords are extracted from the search behavior log according to the user ID, and the search keywords corresponding to the user ID are classified into different search keyword categories according to the user permissions.
  • the related data corresponding to the search keyword category is obtained from the related recommendation set, and the related data is sent to the terminal.
  • user permissions and search keywords can be associated, and related data corresponding to search keywords belonging to the same search keyword category can be sent to user terminals with the same user permissions, thereby achieving association recommendation of related groups and further improving association recommendations. Sex.
  • a step of obtaining related data corresponding to a search keyword category is provided, that is, according to a preset correspondence relationship between a search keyword category and a related recommendation set,
  • the step of acquiring the association data corresponding to the search keyword category in the association recommendation includes the following steps S302 to S310:
  • the server extracts candidate search keywords matching a preset similarity from a database according to the search keywords corresponding to the search keyword category.
  • the server may obtain a search keyword corresponding to the search keyword category by acquiring a correspondence relationship between the search keyword category and the search keyword.
  • a search keyword corresponding to the search keyword category By calculating the similarity between the search keyword and other search keywords in the data, and obtaining other search keywords that match the preset similarity, a post-calculated search keyword is generated.
  • the preset similarity may be set to 0.7, and other search keywords corresponding to the similarity of the search keywords greater than or equal to 0.7 may be selected as candidate search keywords.
  • the obtained search keyword category is "undergraduate institution" as an example.
  • the server obtains the corresponding relationship between the search keyword category and the search keyword, and obtains a search keyword corresponding to the search keyword category “undergraduate institution” and “the distribution of Shanghai male undergraduate institutions”.
  • the other search keywords include "undergraduate institution”, “Shanghai undergraduate institution distribution”, “male undergraduate institution” Institutional distribution "and so on.
  • the server can obtain other search keywords with similarity greater than or equal to 0.7 by calculating similarities between other search keywords and search keywords, and generate candidate search keywords.
  • the server generates a search related candidate vocabulary according to the candidate search keywords matching the preset similarity, and obtains a search result corresponding to the candidate search keywords.
  • the search related candidate thesaurus is composed of a plurality of candidate search keywords matching a preset similarity, and is used to provide corresponding candidate search keywords for different search keywords.
  • the server obtains the corresponding relationship between the candidate search keywords and the search results, and obtains the search results corresponding to the candidate search keywords according to the corresponding relationship between the candidate search keywords and the search results.
  • the obtained search keywords are “the distribution of male undergraduate institutions in Shanghai”, and the candidate search keywords matching the preset similarity are “the distribution of Shanghai undergraduate institutions” and “the distribution of male undergraduate institutions” as examples.
  • the server obtains the search results corresponding to the candidate search keywords "Shanghai undergraduate institution distribution” and "male undergraduate institution distribution”.
  • the server generates a related recommendation set according to the related candidate thesaurus and the search results corresponding to the candidate search keywords.
  • the related recommendation set is composed of related candidate thesaurus and search results corresponding to candidate search keywords, and can be used to provide different users with related recommendation data corresponding to the search request.
  • the related recommendation data is a candidate search keyword corresponding to the search keyword and a search result corresponding to the candidate search keyword.
  • candidate search keywords matching the preset similarity are “Shanghai male borrower” and “distribution of male borrower education” as examples.
  • candidate search keywords “Shanghai male borrower” and “male borrower education distribution”
  • a search result corresponding to a word is used to generate a related recommendation set corresponding to the search keyword "the education distribution of male borrowers in Shanghai”.
  • the server obtains candidate search keywords matching the preset similarity and search results corresponding to the candidate search keywords from the related recommendation set.
  • the server obtains candidate search keywords matching the preset similarity from the related recommendation set according to the search keywords, and obtains the candidate search keywords from the database by obtaining the correspondence between the candidate search keywords and the search results. Corresponding search results.
  • the search keyword obtained is “amount distribution of Shenzhen university student loans” as an example.
  • the candidate search keywords that are obtained from the related recommendation set and meet the preset similarity include: “Shenzhen University Student Loan” and “University Loan Amount Distribution”, etc., and are obtained from the database according to the candidate search keywords that meet the preset similarity.
  • Corresponding search results include "Shenzhen University Student Loan Situation” and “Distribution of University Student Loan Amount in Different Places”. The corresponding search results can be compared and analyzed to obtain the data that most closely matches the search keywords.
  • S310 The server generates association data according to the candidate search keywords that match the preset similarity and the search results corresponding to the candidate search keywords.
  • the association data is composed of candidate search keywords matching the preset similarity and search results corresponding to the candidate search keywords, and is used for sending to the terminal where the corresponding user identifier is located.
  • the server parses the search request sent by the terminal, obtains the user identifier carried by the search request, and obtains the corresponding search keyword from the search behavior log according to the user identifier. Further, the server may obtain candidate search keywords that match the preset similarity from the related recommendation set, wherein the candidate search keywords respectively calculate similarities between other search keywords and search keywords, and obtain a preset value greater than or equal to the preset Similarity of other keywords. The server obtains the corresponding search results from the database according to the candidate search keywords that match the preset similarity, and generates the related data according to the candidate search keywords and the corresponding search results that match the preset similarity, and sends the related data to Corresponds to the terminal where the user ID is located.
  • the server obtains the candidate search keywords that match the preset similarity and the search results corresponding to the candidate search keywords, and generates a related recommendation set according to the related candidate thesaurus and the search results corresponding to the candidate search keywords.
  • the corresponding association data is obtained from the association recommendation set and sent to the corresponding terminal.
  • related data corresponding to different search requests can be sent to the terminal where the corresponding user identifier is located, and targeted related recommendation of the search operation is realized, and the recommended relatedness is improved.
  • a step of obtaining a user's search behavior log from a database according to a search request including:
  • the server parses the search request to obtain the user identification carried in the search request; obtains the historical search operation corresponding to the user identification from the database; and generates a search behavior log of the user according to the historical search operation and the search result corresponding to the historical search operation.
  • the search request carries a user identifier, which is used to indicate that a corresponding user performs a search operation on the terminal and enters a corresponding search phrase.
  • the server can obtain the corresponding user ID and the corresponding search phrase by analyzing the search request.
  • the corresponding historical search operation can be obtained from the database according to the user identification, and the corresponding search result can be obtained from the database according to the historical search operation.
  • the user's search behavior is generated according to the historical search operation and the search result corresponding to the historical search operation. Log.
  • the server may obtain the user's search behavior log from the database according to the search phrase input by the user.
  • the search phrase may be a single word or a phrase or sentence composed of multiple words, or may be Chinese characters or pinyin.
  • the search phrase may be "Shenzhen”, “shenzhen”, and "Shanghai population education distribution”.
  • the server obtains the user identification carried in the search request, and obtains the historical search operation corresponding to the user identification, and generates a search behavior log of the user according to the historical search operation and the search result corresponding to the historical search operation. Therefore, the user ID can be carried for different search requests, and corresponding search behavior logs can be generated, which avoids the problem of using historical search operations and search results corresponding to other user IDs to generate search behavior logs for users, and reduces the error of the search behavior log generation process. To further ensure the accuracy of the search behavior log.
  • a step of extracting a corresponding search keyword from a search behavior log according to a user identifier including:
  • the server obtains the scene corresponding to the user identification and obtains the historical search input corresponding to the scene; obtains the search result corresponding to the historical search input from the search behavior log; and generates the scene corresponding to the scene according to the historical search input and the corresponding search result in the scene Corresponding historical search data; extract search keywords from historical search data.
  • the server obtains the corresponding scene according to the user identifier, and according to the search phrase input by the user, obtains the historical search input corresponding to the search phrase in the corresponding scene from the database, and obtains the search corresponding to the historical search input from the search behavior log. result.
  • the server extracts search keywords from the historical search data in the corresponding scene according to the user identification.
  • the search behavior log of the user includes the historical search operation of the user and the search result corresponding to the historical operation.
  • the search keyword is a keyword that is the same as or similar to any one of the search phrases entered by the user in the terminal.
  • the server may obtain a corresponding user behavior search log according to the search phrase entered by the user, and obtain the corresponding user behavior search log according to the user identifier. In the behavior search log to get the same or similar to the search phrase
  • the server obtains a scene corresponding to the user identifier, obtains a historical search input corresponding to the scene, and obtains a search result corresponding to the historical search input from the search behavior log. Therefore, in the scene corresponding to the user identifier, historical search data corresponding to the scene can be obtained, and corresponding search keywords can be obtained from the historical search data. Before obtaining the search keywords, the scene corresponding to the search request is determined, which reduces multiple Repeated acquisition operations in this scenario further reduce resource consumption.
  • a step of classifying search keywords according to a preset search keyword category according to user permissions including:
  • the server obtains the user identification category corresponding to the user authority; obtains the correspondence between the user identification category and the search keyword, and obtains the search keyword corresponding to the user identification category; the search keywords corresponding to different user identification categories are based on the user
  • the identification categories are divided into different search keyword categories.
  • the server can obtain a user identity category corresponding to the user authority by acquiring a correspondence between the user authority and the user identifier.
  • the server can obtain a search keyword corresponding to the user identification category by acquiring the correspondence between the user identification category and the search keyword.
  • the server can classify the search keywords corresponding to different user identification categories into different search keyword categories by acquiring the association relationship between the search keyword categories and the search keywords.
  • the server can obtain the user identification category corresponding to the user authority and the search keywords corresponding to the user identification category, so that the search keywords corresponding to different user identification categories can be divided into different search keywords according to the user identification category. category. Furthermore, the search keywords corresponding to different user rights are classified, and related data corresponding to search keywords of the same category can be recommended to terminals corresponding to different user IDs of the same user right according to user rights, further strengthening the recommended association. Sex.
  • an association recommendation method is provided, and the method further includes:
  • the server mines input search behaviors and / or click search behaviors performed by users in the network; obtains input search results corresponding to input search behaviors and click search results corresponding to click search behaviors; according to input search behaviors and corresponding inputs performed by users Search results, as well as click search behavior and corresponding click search results, generate a user's search behavior log.
  • the user's search behavior includes input search behavior and click search behavior
  • the corresponding search results include input search results and click search results.
  • the server may input search behavior and corresponding historical input search results according to the user ’s historical input search behavior and corresponding historical input search results
  • the historical click search input and the corresponding historical click search results can generate the corresponding user's search behavior log.
  • the clickable options include: “gender”, “age”, “city”, “location”, “education”, “affiliated college”, “Loans,” “monthly income,” and “consumption preferences”.
  • the server can obtain the corresponding historical click search results from the database according to the user ID corresponding to the user performing the click search and the clicked option, and generate the corresponding user ID according to the options for performing the click search and the corresponding click search results. Search behavior logs.
  • the options clicked by the user include “gender”, "city”, and “loan status”.
  • male is selected in the drop-down item of "gender” option
  • Shenzhen is selected in the drop-down item of "city” option
  • "loan” "Situation” option is selected in the drop-down item, "micro-loans”, comprehensive user click on the option, we can know that the user needs to search the question is "Shenzhen men's micro-loan situation”.
  • the server may obtain the corresponding historical search results from the database according to the user identification and the click search selection, and generate a search behavior log corresponding to the user according to the historical click search options and the corresponding historical search results.
  • the server generates a search log of a user by obtaining a user's input search and corresponding input search result, and a click search and corresponding click search result, so as to generate a personalized search behavior log for the user. , Which is conducive to providing appropriate related recommendations for corresponding users.
  • an association recommendation method is provided, and the method further includes:
  • the server sets the association degree in advance; obtains the candidate search keywords corresponding to the search keywords; sorts the similarity between the candidate search keywords and the search keywords to obtain the first K candidate search keywords; and searches the first K candidate keywords
  • the similarity corresponding to the keywords is compared with the preset correlation to obtain candidate search keywords that match the preset correlation; and the candidate search keywords and corresponding search results that match the preset correlation are sent to the terminal .
  • the correlation degree is used to indicate a preset correlation degree between the search keyword and the candidate search keyword.
  • the server calculates the similarity between the candidate search keywords and the search keywords, sorts the candidate search keywords according to the degree of similarity, and obtains the first K candidate search keywords. Among them, K may be 10.
  • the server calculates the similarity between the candidate search keywords and the search keywords, sorts the candidate search keywords according to the degree of similarity, and obtains the first 10 candidate search keywords.
  • the server compares the similarity corresponding to the first 10 candidate search keywords with a preset correlation degree to obtain a candidate search keyword that matches the preset correlation degree.
  • the correlation degree can be set to 0.8, that is, the server compares the similarity corresponding to the first 10 candidate search keywords with a preset correlation degree of 0.8 to obtain a corresponding degree of similarity greater than or equal to 0.8.
  • Candidate search keywords, and obtain search results corresponding to candidate search keywords with a degree of similarity greater than or equal to 0.8 and generate related data based on the obtained candidate search keywords with a degree of similarity greater than or equal to 0.8, and the corresponding search results, and Send the associated data to the terminal where the corresponding user ID is located.
  • the server sorts the similarity between the candidate search keywords and the search keywords to obtain the first K candidate search keywords, and sets the similarity corresponding to the first K candidate search keywords and the preset
  • the correlations are compared to obtain candidate search keywords matching the preset correlations, and the candidate search keywords and corresponding search results matching the preset correlations are sent to the terminal.
  • further screening of candidate search keywords is achieved, the degree of correlation between the candidate search keywords and the search keywords is ensured, and the relevance of the recommendation is improved.
  • steps in the flowchart of FIG. 2-3 are sequentially displayed in accordance with the directions of the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless explicitly stated in this document, the execution of these steps is not strictly limited, and these steps can be performed in other orders. Moreover, at least a part of the steps in Figure 2-3 may include multiple sub-steps or stages. These sub-steps or stages are not necessarily performed at the same time, but may be performed at different times. These sub-steps or stages The execution order of is not necessarily performed sequentially, but may be performed in turn or alternately with at least a part of another step or a sub-step or stage of another step.
  • an association recommendation device including:
  • the search log acquisition module 402 the search keyword extraction module 404, the classification module 406, and the associated data acquisition module 408, where:
  • the search log obtaining module 402 is configured to receive a search request sent by a terminal, and obtain a search behavior log of a user according to the search request.
  • the search request carries a user ID, and the user ID corresponds to the user authority.
  • the search keyword extraction module 404 is configured to extract a corresponding search keyword from a search behavior log according to a user identifier.
  • a classification module 406 is configured to classify the search keywords according to a preset search keyword category according to user permissions.
  • the association data obtaining module 408 is configured to obtain association data corresponding to the search keyword category from the association recommendation set according to a preset correspondence relationship between the search keyword category and the association recommendation set, and send the association data to the terminal.
  • the server obtains a search behavior log of a user from a database according to a search request sent by a terminal, wherein the search request carries a user identifier, and the user identifier corresponds to a user right.
  • the corresponding search keywords are extracted from the search behavior log according to the user ID, and the search keywords corresponding to the user ID are classified into different search keyword categories according to the user permissions.
  • the related data corresponding to the search keyword category is obtained from the related recommendation set, and the related data is sent to the terminal.
  • user permissions and search keywords can be associated, and related data corresponding to search keywords belonging to the same search keyword category can be sent to user terminals with the same user permissions, thereby achieving association recommendation of related groups and further improving association recommendations. Sex.
  • a search log acquisition module is provided, which is further configured to:
  • the server obtains a user identification carried in the search request, and obtains a historical search operation corresponding to the user identification, and generates a search behavior log of the user according to the historical search operation and the search result corresponding to the historical search operation. Therefore, the user ID can be carried for different search requests, and corresponding search behavior logs can be generated, which avoids the problem of using historical search operations and search results corresponding to other user IDs to generate search behavior logs for users, and reduces the error of the search behavior log generation process To further ensure the accuracy of the search behavior log.
  • a search keyword extraction module is provided, which is further configured to:
  • the server obtains a scene corresponding to a user identifier, obtains a history search input corresponding to the scene, and obtains a search result corresponding to the history search input from a search behavior log. Therefore, in the scene corresponding to the user identifier, historical search data corresponding to the scene can be obtained, and corresponding search keywords can be obtained from the historical search data. Before obtaining the search keywords, the scene corresponding to the search request is determined, which reduces multiple Repeated acquisition operations in this scenario further reduce resource consumption.
  • a classification module is provided, which is further used for:
  • the server obtains the user identification category corresponding to the user authority and the search keywords corresponding to the user identification category, so that the search keywords corresponding to different user identification categories can be divided into different search keys according to the user identification category.
  • Word category Furthermore, the search keywords corresponding to different user rights are classified, and related data corresponding to search keywords of the same category can be recommended to terminals corresponding to different user IDs of the same user right according to user rights, further strengthening the recommended association. Sex.
  • a related data acquisition module is provided, and is further configured to:
  • search keywords corresponding to the search keyword category candidate search keywords matching the preset similarity are extracted from the database; the search related candidate thesaurus is generated based on the candidate search keywords matching the preset similarity, and the key with the candidate search is obtained.
  • Search results corresponding to words generating a related recommendation set according to the related candidate thesaurus and the search results corresponding to the candidate search keywords; obtaining candidate search keywords matching the preset similarity from the related recommendation sets, and corresponding to the candidate search keywords
  • the search results are generated according to the candidate search keywords that match the preset similarity and the search results corresponding to the candidate search keywords, and associated data is generated.
  • the server obtains a candidate search keyword that matches a preset similarity and a search result corresponding to the candidate search keyword, and generates an association according to the related candidate thesaurus and the search result corresponding to the candidate search keyword.
  • the recommendation set when receiving the search request sent by the terminal, obtains the corresponding association data from the association recommendation set and sends it to the corresponding terminal.
  • related data corresponding to different search requests can be sent to the terminal where the corresponding user identifier is located, and targeted related recommendation of the search operation is realized, and the recommended relatedness is improved.
  • an association recommendation device is provided, and the device further includes a search behavior log generating module for:
  • Mining input search behaviors and / or click search behaviors performed by users in the network obtaining input search results corresponding to input search behaviors and click search results corresponding to click search behaviors; according to input search behaviors and corresponding input searches performed by users
  • the results, as well as the click search behavior and the corresponding click search results, generate a user's search behavior log.
  • the server generates a search behavior log corresponding to a user by acquiring a user's input search and corresponding input search results, and a click search and corresponding click search result, thereby realizing a personalized search behavior log for the user
  • the generation is beneficial to provide appropriate related recommendation for corresponding users.
  • an association recommendation device is provided, and the device further includes a sending module, configured to:
  • Relevance is set in advance; candidate search keywords corresponding to the search keywords are obtained; the similarity between the candidate search keywords and the search keywords is sorted to obtain the first K candidate search keywords; the first K candidate search keys are keyed The similarity corresponding to the word is compared with the preset relevance to obtain candidate search keywords that match the preset relevance; and the candidate search keywords and corresponding search results that match the preset relevance are sent to the terminal.
  • the server obtains the first K candidate search keywords by sorting the similarity between the candidate search keywords and the search keywords, and associates the similarity corresponding to the first K candidate search keywords with the preset association
  • the comparison is performed to obtain candidate search keywords matching the preset relevance, and the candidate search keywords matching the preset relevance and the corresponding search results are sent to the terminal.
  • further screening of candidate search keywords is achieved, the degree of correlation between the candidate search keywords and the search keywords is ensured, and the relevance of the recommendation is improved.
  • Each module in the above-mentioned associated recommendation device may be implemented in whole or in part by software, hardware, and a combination thereof.
  • the above-mentioned modules may be embedded in the hardware form or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor calls and performs the operations corresponding to the above modules.
  • a computer device is provided.
  • the computer device may be a server, and its internal structure diagram may be as shown in FIG. 5.
  • the computer device includes a processor, a memory, a network interface, and a database connected through a system bus.
  • the processor of the computer device is used to provide computing and control capabilities.
  • the memory of the computer device includes a non-volatile computer-readable storage medium and an internal memory.
  • the non-volatile computer-readable storage medium stores an operating system, computer-readable instructions, and a database.
  • the internal memory provides an environment for operating systems and computer-readable instructions in a non-volatile computer-readable storage medium.
  • the computer equipment database is used to store associated data.
  • the network interface of the computer device is used to communicate with an external terminal through a network connection.
  • the computer-readable instructions are executed by a processor to implement an association recommendation method.
  • FIG. 5 is only a block diagram of a part of the structure related to the solution of the application, and does not constitute a limitation on the computer equipment to which the solution of the application is applied.
  • the specific computer equipment may be Include more or fewer parts than shown in the figure, or combine certain parts, or have a different arrangement of parts.
  • a computer device includes a memory and one or more processors.
  • the memory stores computer-readable instructions.
  • the method for preprocessing the imbalanced sample data provided in any one of the embodiments of the present application is implemented. A step of.
  • One or more non-volatile computer-readable storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors implement one of the embodiments of the present application Provides steps for pre-processing methods for unbalanced sample data.
  • Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory can include random access memory (RAM) or external cache memory.
  • RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
  • SRAM static RAM
  • DRAM dynamic RAM
  • SDRAM synchronous DRAM
  • DDRSDRAM dual data rate SDRAM
  • ESDRAM enhanced SDRAM
  • SLDRAM synchronous chain Synchlink DRAM
  • Rambus direct RAM
  • DRAM direct memory bus dynamic RAM
  • RDRAM memory bus dynamic RAM

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种基于数据资源的关联推荐方法,包括:接收终端发送的搜索请求,并根据搜索请求获取用户的搜索行为日志。其中,搜索请求携带用户标识,用户标识与用户权限对应。根据用户标识从搜索行为日志中提取对应的搜索关键字,根据用户权限,将搜索关键字按照预设的搜索关键字类别进行分类。根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端。

Description

关联推荐方法、装置、计算机设备和存储介质
相关申请的交叉引用
本申请要求于2018年9月18日提交中国专利局,申请号为2018110889833,申请名称为“关联推荐方法、装置、计算机设备和存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及一种关联推荐方法、装置、计算机设备和存储介质。
背景技术
随着计算机科学技术的不断发展,出现了利用网络获取信息的方式,可以获取网络上的海量信息,并提供了较快的信息获取速度。随之而来的是越来越多的用户通过网络搜索进行各种信息的查询与获取。
传统的搜索技术,当用户通过搜索引擎输入某个查询词后,可以在搜索框下方自动推荐候选词,向用户推荐与用户输入查询语义相关的其他查询。
然而,发明人意识到,传统的搜索技术,当用户通过搜索引擎输入某个查询词后,可以在搜索框下方自动推荐候选词,向用户推荐与用户输入查询语义相关的其他查询。但是,传统的各大搜索引擎的搜索联想功基于各关键词的搜索热度、个人搜索历史,无法结合用户搜索意图为其推送满足其自身需求的服务信息,且不能与用户权限以及关联人群关联起来,推荐关联性较差。
发明内容
根据本申请公开的各种实施例,提供一种关联推荐方法、装置、计算机设备和存储介质。
一种关联推荐方法包括:
接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;所述搜索请求携带用户标识,所述用户标识与用户权限对应;
根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
一种关联推荐装置包括:
搜索日志获取模块,用于接收终端发送的搜索请求,并根据所述搜索请求获取用户的 搜索行为日志;所述搜索请求携带用户标识,所述用户标识与用户权限对应;
搜索关键字提取模块,用于根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
分类模块,用于根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
关联数据获取模块,用于根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
一种计算机设备,包括存储器和一个或多个处理器,所述存储器中储存有计算机可读指令,所述计算机可读指令被所述处理器执行时,使得所述一个或多个处理器执行以下步骤:
接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;所述搜索请求携带用户标识,所述用户标识与用户权限对应;
根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
一个或多个存储有计算机可读指令的非易失性计算机可读存储介质,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器执行以下步骤:
接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;所述搜索请求携带用户标识,所述用户标识与用户权限对应;
根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其它的附图。
图1为根据一个或多个实施例中关联推荐方法的应用场景图。
图2为根据一个或多个实施例中关联推荐方法的流程示意图。
图3为根据一个或多个实施例中获取与搜索关键字类别对应的关联数据的流程示意 图。
图4为根据一个或多个实施例中关联推荐装置的框图。
图5为根据一个或多个实施例中计算机设备的框图。
具体实施方式
为了使本申请的技术方案及优点更加清楚明白,以下结合附图及实施例,对本申请进行进一步详细说明。应当理解,此处描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
本申请提供的关联推荐方法,可以应用于如图1所示的应用环境中。终端102与服务器104通过网络进行通信。服务器104接收终端102发送的搜索请求,并根据搜索请求获取用户的搜索行为日志。搜索请求携带用户标识,用户标识与用户权限对应。服务器104根据用户标识从搜索行为日志中提取对应的搜索关键字,并根据用户权限,将搜索关键字按照预设的搜索关键字类别进行分类。服务器104根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端102。终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机和平板电脑,服务器104可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在其中一个实施例中,如图2所示,提供了一种关联推荐方法,以该方法应用于图1中的服务器为例进行说明,包括以下步骤:
S202,服务器接收终端发送的搜索请求,并根据搜索请求获取用户的搜索行为日志。搜索请求携带用户标识,用户标识与用户权限对应。
具体地,搜索请求携带用户标识,用于表示对应用户在终端执行搜索操作,并输入对应的搜索词组,服务器根据用户输入的搜索词组从数据库中获取用户的搜索行为日志。用户标识与用户权限对应,服务器可根据搜索请求携带的用户标识,获取对应用户所具备的用户权限。
搜索词组可以是单个词语以及多个词语组成的词组或语句,可以是汉字也可以是拼音,比如,搜索词组可以是“深圳”、“shenzhen”以及“上海人群学历分布”等。用户的搜索行为日志包括用户的历史搜索操作和历史操作对应的搜索结果。同时,由于用户的搜索行为包括输入搜索行为和点击搜索行为,因此,对应的搜索结果包括输入搜索结果和点击搜索结果,可根据用户的历史输入搜索行为和对应历史输入搜索结果,以及用户的历史点击搜索输入和对应的历史点击搜索结果,可生成对应的用户的搜索行为日志。
进一步地,服务器接收终端的搜索请求,并获取搜索请求携带的用户标识,根据用户标识和用户权限之间的对应关系,获取与用户标识对应的用户权限。根据用户输入的搜索词组,从数据库中获取与用户标识对应的历史搜索操作和历史搜索操作对应的搜索结果,同时还可根据该用户的用户权限,获取同一用户权限的不同用户的历史搜索操作和对应的搜索结果。
以用户在终端进行输入搜索为例,用户输入的搜索词组为“上海男性借款人的学历如何分布”,服务器接收终端发送的搜索请求,在数据库中获取该用户标识对应的历史搜索操作,包括“上海”、“男性借款人”、“学历分布”、“上海男性学历分布”以及“借款人学历分布”等,多项历史搜索操作,并从数据库中获取与历史搜索操作对应的搜索结果。根据多项历史操作和对应的搜索结果,生成与用户输入的“上海男性借款人的学历分布如何”对应的用户的搜索行为日志。
S204,服务器根据用户标识从搜索行为日志中提取对应的搜索关键字。
具体地,服务器根据用户标识获取对应的场景,并根据用户输入的搜索词组,从数据库中获取对应场景下与搜索词组对应的历史搜索输入,并从搜索行为日志中获取与历史搜索输入对应的搜索结果。根据场景下与用户输入的搜索词组对应的历史搜索输入,以及与历史搜索输入对应的搜索结果,可生产对应场景下的历史搜索数据。服务器根据用户标识从对应场景下的历史搜索数据中,提取搜索关键字。
用户的搜索行为日志包括用户的历史搜索操作和历史操作对应的搜索结果。搜索关键字为与用户在终端输入的搜索词组中任一搜索词相同或相近的关键字,服务器可根据用户输入的搜索词组,并根据用户标识获取得到对应的用户的行为搜索日志,从所获得的行为搜索日志中获取与搜索词组相同或相近的搜索关键字。
进一步地,以用户进行输入搜索为例。用户输入的搜索词组为“深圳小贷借款人的学历分布如何”,对应的用户搜索行为日志中的历史搜索操作,包括“深圳”、“小贷借款人”、“学历分布”、“深圳小贷借款人”以及“小贷借款人的学历分布”等。根据用户标识,可从对应的用户搜索行为日志中获取得到的搜索关键字,可以是:“深圳小贷借款人”、“小贷借款人学历分布”以及“深圳”等。
S206,服务器根据用户权限,将搜索关键字按照预设的搜索关键字类别进行分类。
具体地,服务器通过获取用户权限和用户标识之间的对应关系,可获取与用户权限对应的用户标识类别。服务器通过获取用户标识类别和搜索关键字之间的对应关系,可获取与用户标识类别对应的搜索关键字。服务器通过获取搜索关键字类别和搜索关键字之间的关联关系,可将与不同用户标识类别对应的搜索关键字,归类至不同的搜索关键字类别。
用户权限表示在对应的终端上用户具备的权限,包括搜索、添加、删除以及更新等,不同用户可具备同一用户权限。比如,用户A在终端上具备搜索和更新的权限,用户B在终端上具备搜索的权限,用户C在终端上具备更新的权限。用户标识类别表示相同用户权限的不同用户,所具备的标识对应的类别,也就是说,用户A和用户B均具备搜索权限,但用户B不具备更新权限,因此用户A和用户B对应不同的用户标识类别。
进一步地,根据预设的搜索关键字类别和搜索关键字类别之间的关联关系,可将与不同用户标识类别对应的搜索关键字,归类至不同的搜索关键字类别。比如,获取到的用户A进行输入的搜索关键字为“深圳大学生贷款数额分布”,则根据用户A输入的搜索关键字,可获取到对应的搜索关键字类别为“贷款”。根据用户A对应的用户标识类别,将所 获取到的用户A输入的搜索关键字“深圳大学生贷款数额分布”,归类至为“贷款”对应的搜索关键字类别。
S208,服务器根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端。
具体地,服务器根据搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字。根据符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与候选搜索关键字对应的搜索结果。根据关联候选词库和与候选搜索关键字对应的搜索结果,生成关联推荐集,并从关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果。根据符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果,生成关联数据。
服务器通过计算搜索关键字和数据库中的其他搜索关键字之间的相似度,并得到符合预设的相似度的候选搜索关键字,根据符合预设的相似度的候选搜索关键字生成搜索关联候选词库。在本实施例中,可将预设的相似度设置为0.7,即可将与搜索关键字的相似度为0.7对应的其他搜索关键字挑选出来,作为候选搜索关键字,根据大于等于预设相似度0.7的候选搜索关键字,生成搜索关联候选词库。
服务器可根据获取到的与候选搜索关键字对应的搜索结果,以及根据候选搜索关键字生成的搜索关联候选词库,生成关联推荐集。进一步地,服务器根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据。关联数据由符合预设相似度的候选搜索关键字和与候选搜索关键字对应的搜索结果组成。
上述关联推荐方法中,服务器通过根据终端发送的搜索请求从数据库中获取用户的搜索行为日志,搜索请求携带用户标识,用户标识与用户权限对应。根据用户标识从搜索行为日志中提取对应的搜索关键字,并根据用户权限,将用户标识对应的搜索关键字分为不同搜索关键字类别。根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端。从而可将用户权限和搜索关键字关联起来,将属于同一搜索关键字类别的搜索关键字对应的关联数据,发送至相同用户权限的用户终端,实现了关联人群的关联推荐,进一步提高了关联推荐性。
在其中一个实施例中,如图3所示,提供了一种获取与搜索关键字类别对应的关联数据的步骤,即根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据的步骤,包括以下步骤S302至S310:
S302,服务器根据搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字。
具体地,服务器通过获取搜索关键字类别和搜索关键字之间的对应关系,可获取与搜索关键字类别对应的搜索关键字。通过计算搜索关键字和数据中其他搜索关键字之间的相似度,并且获取符合预设相似度的其他搜索关键字,生成后算搜索关键字。在本实施例 中,可将预设的相似度设置为0.7,即可将与搜索关键字的相似度大于等于0.7对应的其他搜索关键字挑选出来,作为候选搜索关键字。
进一步地,以获取得到的搜索关键字类别为“本科机构”为例。服务器通过获取搜索关键字类别和搜索关键字之间的对应关系,获取与搜索关键字类别为“本科机构”对应的为“上海男性本科的机构分布”的搜索关键字。通过计算数据库中其他搜算关键字和搜索关键字“上海男性本科的机构分布”之间的相似度,其中,其他搜算关键字包括“本科机构”、“上海本科机构分布”、“男性本科机构分布”等。服务器通过分别计算其他搜算关键字和搜索关键字之间的相似度,可获取相似度大于等于0.7的其他搜算关键字,生成候选搜索关键字。
S304,服务器根据符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与候选搜索关键字对应的搜索结果。
具体地,搜索关联候选词库由多个符合预设相似度的候选搜索关键字组成,用于针对不同的搜索关键字提供对应的候选搜索关键字。服务器通过获取候选搜索关键字和搜索结果之间的对应关系,并根据候选搜索关键字和搜索结果之间的对应关系,获取与候选搜索关键字对应的搜索结果。
进一步地,以获取到的搜索关键字为“上海男性本科的机构分布”,符合预设相似度的候选搜索关键字为“上海本科机构分布”以及“男性本科机构分布”为例。服务器根据候选搜索关键字“上海本科机构分布”和“男性本科机构分布”,获取与之对应的搜索结果。
S306,服务器根据关联候选词库和与候选搜索关键字对应的搜索结果,生成关联推荐集。
具体地,关联推荐集由关联候选词库和与候选搜索关键字对应的搜索结果组成,可用于向不同用户提供与搜索请求对应的关联推荐数据。其中,关联推荐数据为与搜索关键字对应的候选搜索关键字,以及与候选搜索关键字对应的搜索结果。
进一步地,以获取到的搜索关键字为“上海男性借款人的学历分布”,符合预设相似度的候选搜索关键字为“上海男性借款人”以及“男性借款人学历分布”为例。根据候选搜索关键字“上海男性借款人”以及“男性借款人学历分布”,生成对应的关联候选词库,并获取与候选搜索关键字对应的搜索结果,根据关联候选词库和与候选搜索关键字对应的搜索结果,生成与搜索关键字“上海男性借款人的学历分布”对应的关联推荐集。
S308,服务器从关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果。
具体地,服务器根据搜索关键字,从关联推荐集中获取符合预设相似度的候选搜索关键字,并通过获取候选搜索关键字和搜索结果之间的对应关系,从数据库中获取与候选搜索关键字对应的搜索结果。
进一步地,以获取到的搜索关键字为“深圳大学生贷款的数额分布”为例。从关联 推荐集中获取到的,符合预设相似度的候选搜索关键字包括:“深圳大学生贷款”以及“大学生贷款数额分布”等,根据符合预设相似度的候选搜索关键字,从数据库中获取对应的搜索结果,包括“深圳大学生贷款情况”以及“各地大学生贷款数额分布”,可将对应的搜索结果进行对比分析,获得与搜索关键字最符合的数据。
S310,服务器根据符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果,生成关联数据。
具体地,关联数据由符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果组成,用于发送至对应用户标识所在的终端。
进一步地,服务器通过解析终端发送的搜索请求,并获取搜索请求携带的用户标识,根据用户标识从搜索行为日志中获取对应的搜索关键字。进而,服务器可从关联推荐集中获取符合预设相似度的候选搜索关键字,其中,候选搜索关键字通过分别计算其他搜算关键字和搜索关键字之间的相似度,并获取大于等于预设相似度的其他关键字得到。服务器根据符合预设相似度的候选搜索关键字,从数据库中获取对应的搜索结果,并根据符合预设相似度的候选搜索关键字和对应的搜索结果,生成关联数据,并将关联数据发送至对应用户标识所在终端。
上述步骤,服务器通过获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果,并根据关联候选词库和与候选搜索关键字对应的搜索结果,生成关联推荐集,当接收到终端发送的搜索请求时,从关联推荐集中获取对应的关联数据,发送至对应的终端。从而可将不同搜索请求对应的关联数据发送至对应的用户标识所在的终端,实现了搜索操作的针对性关联推荐,提高了推荐的关联性。
在其中一个实施例中,提供了一种根据搜索请求从数据库中获取用户的搜索行为日志的步骤,包括:
服务器解析搜索请求,获取搜索请求携带的用户标识;从数据库中获取与用户标识对应的历史搜索操作;根据历史搜索操作和与历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
具体地,搜索请求携带用户标识,用于表示对应用户在终端执行搜索操作,并输入对应的搜索词组。服务器通过解析搜索请求,可获取对应的用户标识和对应的搜索词组。进而可根据用户标识,从数据库中获取对应的历史搜索操作,并根据历史搜索操作,从数据库中获取对应的搜索结果,根据历史搜索操作和与历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
进一步地,服务器可根据用户输入的搜索词组从数据库中获取用户的搜索行为日志。其中,搜索词组可以是单个词语以及多个词语组成的词组或语句,可以是汉字也可以是拼音,比如,搜索词组可以是“深圳”、“shenzhen”以及“上海人群学历分布”等。
上述步骤,服务器通过获取搜索请求携带的用户标识,并获取与用户标识对应的历史搜索操作,根据历史搜索操作和与历史搜索操作对应的搜索结果,生成用户的搜索行为 日志。从而可针对不同搜索请求携带用户标识,生成对应的搜索行为日志,避免出现利用其他用户标识对应的历史搜索操作和搜索结果,生成用户的搜索行为日志的问题,降低搜索行为日志生成过程的误差性,进一步保证了搜索行为日志的准确度。
在其中一个实施例中,提供了一种根据用户标识从搜索行为日志中提取对应的搜索关键字的步骤,包括:
服务器获取与用户标识对应的场景,并获取与场景对应的历史搜索输入;从搜索行为日志中获取与历史搜索输入对应的搜索结果;根据场景下的历史搜索输入和对应的搜索结果,生成与场景对应的历史搜索数据;从历史搜索数据中提取搜索关键字。
具体地,服务器根据用户标识获取对应的场景,并根据用户输入的搜索词组,从数据库中获取对应场景下与搜索词组对应的历史搜索输入,并从搜索行为日志中获取与历史搜索输入对应的搜索结果。根据场景下与用户输入的搜索词组对应的历史搜索输入,以及与历史搜索输入对应的搜索结果,可生产对应场景下的历史搜索数据。服务器根据用户标识从对应场景下的历史搜索数据中,提取搜索关键字。
其中,用户的搜索行为日志包括用户的历史搜索操作和历史操作对应的搜索结果。搜索关键字为与用户在终端输入的搜索词组中任一搜索词相同或相近的关键字,服务器可根据用户输入的搜索词组,并根据用户标识获取得到对应的用户的行为搜索日志,从所获得的行为搜索日志中获取与搜索词组相同或相
上述步骤,服务器通过获取与用户标识对应的场景,获取与场景对应的历史搜索输入,并从搜索行为日志中获取与历史搜索输入对应的搜索结果。从而可在与用户标识对应的场景下,获取与场景对应的历史搜索数据,并从历史搜索数据中获取对应的搜索关键字,在获取搜索关键字之前,确定搜索请求对应的场景,减少了多个场景下的重复获取操作,进一步降低了资源消耗。
在其中一个实施例中,提供了一种根据用户权限,将搜索关键字按照预设的搜索关键字类别进行分类的步骤,包括:
服务器获取用户权限对应的用户标识类别;获取用户标识类别和搜索关键字之间的对应关系,并获取与用户标识类别对应的搜索关键字;将与不同用户标识类别对应的搜索关键字,按照用户标识类别分为不同搜索关键字类别。
具体地,服务器通过获取用户权限和用户标识之间的对应关系,可获取与用户权限对应的用户标识类别。服务器通过获取用户标识类别和搜索关键字之间的对应关系,可获取与用户标识类别对应的搜索关键字。服务器通过获取搜索关键字类别和搜索关键字之间的关联关系,可将与不同用户标识类别对应的搜索关键字,归类至不同的搜索关键字类别。
上述步骤,服务器通过获取用户权限对应的用户标识类别,并获取与用户标识类别对应的搜索关键字,从而可将与不同用户标识类别对应的搜索关键字,按照用户标识类别分为不同搜索关键字类别。进而实现了将不同用户权限对应的搜索关键字进行分类,可根据用户权限将相同类别的搜索关键字对应的关联数据,推荐给同一用户权限的不同用户标 识对应的终端,进一步加强了推荐的关联性。
在其中一个实施例中,提供了一种关联推荐方法,该方法还包括:
服务器挖掘网络中用户执行的输入搜索行为和/或点击搜索行为;获取与输入搜索行为对应的输入搜索结果,及与点击搜索行为对应的点击搜索结果;根据用户执行的输入搜索行为和对应的输入搜索结果,以及点击搜索行为和对应的点击搜索结果,生成用户的搜索行为日志。
具体地,用户的搜索行为包括输入搜索行为和点击搜索行为,与之对应的搜索结果包括输入搜索结果和点击搜索结果,服务器可根据用户的历史输入搜索行为和对应历史输入搜索结果,以及用户的历史点击搜索输入和对应的历史点击搜索结果,可生成对应的用户的搜索行为日志。
进一步地,以用户在终端进行点击搜索为例,其中,可供点击的选项包括:“性别”、“年龄”、“所在城市”、“所在机构”、“学历”、“所属院校”、“贷款情况”、“月收入”以及“消费偏好”等。服务器可根据进行点击搜索的用户对应的用户标识,以及所点击的选项,从数据库中获取对应的历史点击搜索结果,根据进行点击搜索点击的选项,和对应的点击搜索结果,生成对应用户标识的搜索行为日志。
比如,用户点击的选项包括“性别”、“所在城市”以及“贷款情况”,其中,“性别”选项的下拉项目中选择了男性,“所在城市”选项的下拉项目中选择了深圳,“贷款情况”选项的下拉项目中选择了“小额贷款”,综合用户点击的选项,可得知用户需要进行搜索的问题为“深圳男性的小额贷款情况”。服务器可根据用户标识和点击搜索的选择,从数据库中获取对应的历史搜索结果,并根据历史点击搜索选项和对应历史搜索结果,生成对应用户的搜索行为日志。
上述关联推荐方法,服务器通过获取用户的输入搜索和对应的输入搜索结果,以及点击搜索和对应的点击搜索结果,生成对应用户的搜索行为日志,从而可实现针对用户的个性化搜索行为日志的生成,有利于为对应用户提供合适的关联推荐。
在其中一个实施例中,提供了一种关联推荐方法,该方法还包括:
服务器预先设置关联度;获取与搜索关键字对应的候选搜索关键字;对候选搜索关键字和搜索关键字之间的相似度进行排序,获取前K个候选搜索关键字;将前K个候选搜索关键字对应的相似度和预设的关联度进行比对,获取符合预设的关联度的候选搜索关键字;将符合预设的关联度的候选搜索关键字和对应的搜索结果,发送至终端。
具体地,关联度用于表示搜索关键字和候选搜索关键字之间的预设关联程度。服务器通过计算候选搜索关键字和搜索关键字之间的相似度,按照相似度的大小,对候选搜索关键字进行排序,并获取前K个候选搜索关键字。其中,K可取10,服务器通过计算候选搜索关键字和搜索关键字之间的相似度,按照相似度的大小,对候选搜索关键字进行排序,并获取前10个候选搜索关键字。
进一步地,服务器将前10个候选搜索关键字对应的相似度和预设的关联度进行比对, 获取符合预设的关联度的候选搜索关键字。其中,可将关联度设置为0.8,也就是说,服务器通过将前10个候选搜索关键字对应的相似度,和预设的关联度0.8进行比对,获取大于或等于0.8的相似度对应的候选搜索关键字,并获取相似度大于或等于0.8的候选搜索关键字对应的搜索结果,根据所获得的相似度大于或等于0.8的候选搜索关键字,以及对应的搜索结果,生成关联数据,并将关联数据发送至对应用户标识所在的终端。
上述关联推荐方法,服务器通过对候选搜索关键字和搜索关键字之间的相似度进行排序,获取前K个候选搜索关键字,并将前K个候选搜索关键字对应的相似度和预设的关联度进行比对,进而可获取符合预设的关联度的候选搜索关键字,并将符合预设的关联度的候选搜索关键字和对应的搜索结果,发送至终端。从而实现了对候选搜索关键字的进一步筛选,保证了候选搜索关键字和搜索关键字之间的关联度,提高了推荐的关联性。
应该理解的是,虽然图2-3的流程图中的各个步骤按照箭头的指示依次显示,但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明,这些步骤的执行并没有严格的顺序限制,这些步骤可以以其它的顺序执行。而且,图2-3中的至少一部分步骤可以包括多个子步骤或者多个阶段,这些子步骤或者阶段并不必然是在同一时刻执行完成,而是可以在不同的时刻执行,这些子步骤或者阶段的执行顺序也不必然是依次进行,而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。
在其中一个实施例中,如图4所示,提供了一种关联推荐装置,包括:
搜索日志获取模块402、搜索关键字提取模块404、分类模块406以及关联数据获取模块408,其中:
搜索日志获取模块402,用于接收终端发送的搜索请求,并根据搜索请求获取用户的搜索行为日志。其中,搜索请求携带用户标识,用户标识与用户权限对应。
搜索关键字提取模块404,用于根据用户标识从搜索行为日志中提取对应的搜索关键字。
分类模块406,用于根据用户权限,将搜索关键字按照预设的搜索关键字类别进行分类。
关联数据获取模块408,用于根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端。
上述关联推荐装置,服务器通过根据终端发送的搜索请求从数据库中获取用户的搜索行为日志,其中,搜索请求携带用户标识,用户标识与用户权限对应。根据用户标识从搜索行为日志中提取对应的搜索关键字,并根据用户权限,将用户标识对应的搜索关键字分为不同搜索关键字类别。根据预设的搜索关键字类别和关联推荐集之间的对应关系,从关联推荐集中获取与搜索关键字类别对应的关联数据,并将关联数据发送至终端。从而可将用户权限和搜索关键字关联起来,将属于同一搜索关键字类别的搜索关键字对应的关联数 据,发送至相同用户权限的用户终端,实现了关联人群的关联推荐,进一步提高了关联推荐性。
在其中一个实施例中,提供了一种搜索日志获取模块,还用于:
解析搜索请求,获取搜索请求携带的用户标识;从数据库中获取与用户标识对应的历史搜索操作;根据历史搜索操作和与历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
上述搜索日志获取模块,服务器通过获取搜索请求携带的用户标识,并获取与用户标识对应的历史搜索操作,根据历史搜索操作和与历史搜索操作对应的搜索结果,生成用户的搜索行为日志。从而可针对不同搜索请求携带用户标识,生成对应的搜索行为日志,避免出现利用其他用户标识对应的历史搜索操作和搜索结果,生成用户的搜索行为日志的问题,降低搜索行为日志生成过程的误差性,进一步保证了搜索行为日志的准确度。
在其中一个实施例中,提供了一种搜索关键字提取模块,还用于:
获取与用户标识对应的场景,并获取与场景对应的历史搜索输入;从搜索行为日志中获取与历史搜索输入对应的搜索结果;根据场景下的历史搜索输入和对应的搜索结果,生成与场景对应的历史搜索数据;从历史搜索数据中提取搜索关键字。
上述搜索关键字提取模块,服务器通过获取与用户标识对应的场景,获取与场景对应的历史搜索输入,并从搜索行为日志中获取与历史搜索输入对应的搜索结果。从而可在与用户标识对应的场景下,获取与场景对应的历史搜索数据,并从历史搜索数据中获取对应的搜索关键字,在获取搜索关键字之前,确定搜索请求对应的场景,减少了多个场景下的重复获取操作,进一步降低了资源消耗。
在其中一个实施例中,提供了一种分类模块,还用于:
获取用户权限对应的用户标识类别;获取用户标识类别和搜索关键字之间的对应关系,并获取与用户标识类别对应的搜索关键字;将与不同用户标识类别对应的搜索关键字,按照用户标识类别分为不同搜索关键字类别。
上述分类模块,服务器通过获取用户权限对应的用户标识类别,并获取与用户标识类别对应的搜索关键字,从而可将与不同用户标识类别对应的搜索关键字,按照用户标识类别分为不同搜索关键字类别。进而实现了将不同用户权限对应的搜索关键字进行分类,可根据用户权限将相同类别的搜索关键字对应的关联数据,推荐给同一用户权限的不同用户标识对应的终端,进一步加强了推荐的关联性。
在其中一个实施例中,提供了一种关联数据获取模块,还用于:
根据搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字;根据符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与候选搜索关键字对应的搜索结果;根据关联候选词库和与候选搜索关键字对应的搜索结果,生成关联推荐集;从关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键 字对应的搜索结果;根据符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果,生成关联数据。
上述关联数据获取模块,服务器通过获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果,并根据关联候选词库和与候选搜索关键字对应的搜索结果,生成关联推荐集,当接收到终端发送的搜索请求时,从关联推荐集中获取对应的关联数据,发送至对应的终端。从而可将不同搜索请求对应的关联数据发送至对应的用户标识所在的终端,实现了搜索操作的针对性关联推荐,提高了推荐的关联性。
在其中一个实施例中,提供了一种关联推荐装置,该装置还包括搜索行为日志生成模块,用于:
挖掘网络中用户执行的输入搜索行为和/或点击搜索行为;获取与输入搜索行为对应的输入搜索结果,及与点击搜索行为对应的点击搜索结果;根据用户执行的输入搜索行为和对应的输入搜索结果,以及点击搜索行为和对应的点击搜索结果,生成用户的搜索行为日志。
上述搜索行为日志生成模块,服务器通过获取用户的输入搜索和对应的输入搜索结果,以及点击搜索和对应的点击搜索结果,生成对应用户的搜索行为日志,从而可实现针对用户的个性化搜索行为日志的生成,有利于为对应用户提供合适的关联推荐。
在其中一个实施例中,提供了一种关联推荐装置,该装置还包括发送模块,用于:
预先设置关联度;获取与搜索关键字对应的候选搜索关键字;对候选搜索关键字和搜索关键字之间的相似度进行排序,获取前K个候选搜索关键字;将前K个候选搜索关键字对应的相似度和预设的关联度进行比对,获取符合预设的关联度的候选搜索关键字;将符合预设的关联度的候选搜索关键字和对应的搜索结果,发送至终端。
上述发送模块,服务器通过对候选搜索关键字和搜索关键字之间的相似度进行排序,获取前K个候选搜索关键字,并将前K个候选搜索关键字对应的相似度和预设的关联度进行比对,进而可获取符合预设的关联度的候选搜索关键字,并将符合预设的关联度的候选搜索关键字和对应的搜索结果,发送至终端。从而实现了对候选搜索关键字的进一步筛选,保证了候选搜索关键字和搜索关键字之间的关联度,提高了推荐的关联性。
关于关联推荐装置的具体限定可以参见上文中对于关联推荐方法的限定,在此不再赘述。上述关联推荐装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中,也可以以软件形式存储于计算机设备中的存储器中,以便于处理器调用执行以上各个模块对应的操作。
在一个实施例中,提供了一种计算机设备,该计算机设备可以是服务器,其内部结构图可以如图5所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口和数据库。其中,该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性计算机可读存储介质、内存储器。该非易失性计算机可读存储介质存储有操作系统、计算机可读指令和数据库。该内存储器为非易失性计算机可读存储介质中的操作 系统和计算机可读指令的运行提供环境。该计算机设备的数据库用于存储关联数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种关联推荐方法。
本领域技术人员可以理解,图5中示出的结构,仅仅是与本申请方案相关的部分结构的框图,并不构成对本申请方案所应用于其上的计算机设备的限定,具体的计算机设备可以包括比图中所示更多或更少的部件,或者组合某些部件,或者具有不同的部件布置。
一种计算机设备,包括存储器和一个或多个处理器,存储器中存储有计算机可读指令,计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的不平衡样本数据预处理方法的步骤。
一个或多个存储有计算机可读指令的非易失性计算机可读存储介质,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器实现本申请任意一个实施例中提供的不平衡样本数据预处理方法的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机可读指令来指令相关的硬件来完成,所述的计算机可读指令可存储于一非易失性计算机可读取存储介质中,该计算机可读指令在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
以上实施例的各技术特征可以进行任意的组合,为使描述简洁,未对上述实施例中的各个技术特征所有可能的组合都进行描述,然而,只要这些技术特征的组合不存在矛盾,都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式,其描述较为具体和详细,但并不能因此而理解为对发明专利范围的限制。应当指出的是,对于本领域的普通技术人员来说,在不脱离本申请构思的前提下,还可以做出若干变形和改进,这些都属于本申请的保护范围。因此,本申请专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种关联推荐方法,包括:
    接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;其中,所述搜索请求携带用户标识,所述用户标识与用户权限对应;
    根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
    根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
    根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
  2. 根据权利要求1所述的方法,其特征在于,所述根据所述搜索请求从数据库中获取用户的搜索行为日志,包括:
    解析所述搜索请求,获取所述搜索请求携带的用户标识;
    从数据库中获取与所述用户标识对应的历史搜索操作;及
    根据所述历史搜索操作和与所述历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
  3. 根据权利要求1所述的方法,其特征在于,所述根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字,包括:
    获取与用户标识对应的场景,并获取与所述场景对应的历史搜索输入;
    根据所述场景下的历史搜索输入和对应的搜索结果,生成与所述场景对应的历史搜索数据;及
    从所述历史搜索数据中提取搜索关键字。
  4. 根据权利要求1至3任意一项所述的方法,其特征在于,所述根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类,包括:
    获取所述用户权限对应的用户标识类别;
    获取所述用户标识类别和所述搜索关键字之间的对应关系,并获取与所述用户标识类别对应的搜索关键字;及
    将与不同用户标识类别对应的搜索关键字,按照所述用户标识类别分为不同搜索关键字类别。
  5. 根据权利要求1至3任意一项所述的方法,其特征在于,所述根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,包括:
    根据所述搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字;
    根据所述符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与所述候选搜索关键字对应的搜索结果;
    根据所述关联候选词库和与所述候选搜索关键字对应的搜索结果,生成关联推荐集;
    从所述关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果;及
    根据所述符合预设相似度的候选搜索关键字,和所述与候选搜索关键字对应的搜索结果,生成关联数据。
  6. 根据权利要求1至3任意一项所述的方法,其特征在于,还包括:
    挖掘网络中用户执行的输入搜索行为和/或点击搜索行为;
    获取与所述输入搜索行为对应的输入搜索结果,及与所述点击搜索行为对应的点击搜索结果;及
    根据用户执行的输入搜索行为和对应的输入搜索结果,以及点击搜索行为和对应的点击搜索结果,生成用户的搜索行为日志。
  7. 根据权利要求1至3任一项所述的方法,其特征在于,还包括:
    预先设置关联度;
    获取与所述搜索关键字对应的候选搜索关键字;
    对所述候选搜索关键字和搜索关键字之间的相似度进行排序,获取前K个候选搜索关键字;
    将所述前K个候选搜索关键字对应的相似度和预设的关联度进行比对,获取符合预设的关联度的候选搜索关键字;及
    将所述符合预设的关联度的候选搜索关键字和对应的搜索结果,发送至终端。
  8. 一种关联推荐装置,包括:
    搜索日志获取模块,用于接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;其中,所述搜索请求携带用户标识,所述用户标识与用户权限对应;
    搜索关键字提取模块,用于根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
    分类模块,用于根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
    关联数据获取模块,用于根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
  9. 根据权利要求8所述的装置,其特征在于,所述搜索日志获取模块还用于
    解析所述搜索请求,获取所述搜索请求携带的用户标识;
    从数据库中获取与所述用户标识对应的历史搜索操作;及
    根据所述历史搜索操作和与所述历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
  10. 一种计算机设备,包括存储器及一个或多个处理器,所述存储器中储存有计算机可读指令,所述计算机可读指令被所述一个或多个处理器执行时,使得所述一个或多个处 理器执行以下步骤:
    接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;其中,所述搜索请求携带用户标识,所述用户标识与用户权限对应;
    根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
    根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
    根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
  11. 根据权利要求10所述的计算机设备,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    解析所述搜索请求,获取所述搜索请求携带的用户标识;
    从数据库中获取与所述用户标识对应的历史搜索操作;及
    根据所述历史搜索操作和与所述历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
  12. 根据权利要求10所述的计算机设备,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    获取与用户标识对应的场景,并获取与所述场景对应的历史搜索输入;
    根据所述场景下的历史搜索输入和对应的搜索结果,生成与所述场景对应的历史搜索数据;及
    从所述历史搜索数据中提取搜索关键字。
  13. 根据权利要求10至12任一项所述的计算机设备,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    获取所述用户权限对应的用户标识类别;
    获取所述用户标识类别和所述搜索关键字之间的对应关系,并获取与所述用户标识类别对应的搜索关键字;及
    将与不同用户标识类别对应的搜索关键字,按照所述用户标识类别分为不同搜索关键字类别。
  14. 根据权利要求10至12任一项所述的计算机设备,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    根据所述搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字;
    根据所述符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与所述候选搜索关键字对应的搜索结果;
    根据所述关联候选词库和与所述候选搜索关键字对应的搜索结果,生成关联推荐集;
    从所述关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果;及
    根据所述符合预设相似度的候选搜索关键字,和所述与候选搜索关键字对应的搜索结果,生成关联数据。
  15. 根据权利要求10至12任一项所述的计算机设备,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    挖掘网络中用户执行的输入搜索行为和/或点击搜索行为;
    获取与所述输入搜索行为对应的输入搜索结果,及与所述点击搜索行为对应的点击搜索结果;及
    根据用户执行的输入搜索行为和对应的输入搜索结果,以及点击搜索行为和对应的点击搜索结果,生成用户的搜索行为日志。
  16. 一个或多个存储有计算机可读指令的非易失性计算机可读存储介质,所述计算机可读指令被一个或多个处理器执行时,使得所述一个或多个处理器执行以下步骤:
    接收终端发送的搜索请求,并根据所述搜索请求获取用户的搜索行为日志;其中,所述搜索请求携带用户标识,所述用户标识与用户权限对应;
    根据所述用户标识从所述搜索行为日志中提取对应的搜索关键字;
    根据所述用户权限,将所述搜索关键字按照预设的搜索关键字类别进行分类;及
    根据预设的搜索关键字类别和关联推荐集之间的对应关系,从所述关联推荐集中获取与所述搜索关键字类别对应的关联数据,并将所述关联数据发送至终端。
  17. 根据权利要求16所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    解析所述搜索请求,获取所述搜索请求携带的用户标识;
    从数据库中获取与所述用户标识对应的历史搜索操作;及
    根据所述历史搜索操作和与所述历史搜索操作对应的搜索结果,生成用户的搜索行为日志。
  18. 根据权利要求16所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    获取与用户标识对应的场景,并获取与所述场景对应的历史搜索输入;
    根据所述场景下的历史搜索输入和对应的搜索结果,生成与所述场景对应的历史搜索数据;及
    从所述历史搜索数据中提取搜索关键字。
  19. 根据权利要求16至18任一项所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    获取所述用户权限对应的用户标识类别;
    获取所述用户标识类别和所述搜索关键字之间的对应关系,并获取与所述用户标识类别对应的搜索关键字;及
    将与不同用户标识类别对应的搜索关键字,按照所述用户标识类别分为不同搜索关键 字类别。
  20. 根据权利要求16至18任一项所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    根据所述搜索关键字类别对应的搜索关键字,从数据库中提取符合预设相似度的候选搜索关键字;
    根据所述符合预设相似度的候选搜索关键字生成搜索关联候选词库,并获取与所述候选搜索关键字对应的搜索结果;
    根据所述关联候选词库和与所述候选搜索关键字对应的搜索结果,生成关联推荐集;
    从所述关联推荐集中获取符合预设相似度的候选搜索关键字,和与候选搜索关键字对应的搜索结果;及
    根据所述符合预设相似度的候选搜索关键字,和所述与候选搜索关键字对应的搜索结果,生成关联数据。
PCT/CN2019/071180 2018-09-18 2019-01-10 关联推荐方法、装置、计算机设备和存储介质 Ceased WO2020057022A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
KR1020217003560A KR20210024173A (ko) 2018-09-18 2019-01-10 연관추천방법, 장치, 컴퓨터 장비 및 저장매체
SG11202101614VA SG11202101614VA (en) 2018-09-18 2019-01-10 Association recommendation method and device, computer equipment and storage medium
JP2021504196A JP7073576B2 (ja) 2018-09-18 2019-01-10 関連付け推薦方法、装置、コンピュータ機器及び記憶媒体
EP19863131.9A EP3855324A4 (en) 2018-09-18 2019-01-10 ASSOCIATIVE RECOMMENDATION PROCESS AND APPARATUS, ASSOCIATED COMPUTER DEVICE AND STORAGE MEDIA

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811088983.3A CN109446407A (zh) 2018-09-18 2018-09-18 关联推荐方法、装置、计算机设备和存储介质
CN201811088983.3 2018-09-18

Publications (1)

Publication Number Publication Date
WO2020057022A1 true WO2020057022A1 (zh) 2020-03-26

Family

ID=65530407

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/071180 Ceased WO2020057022A1 (zh) 2018-09-18 2019-01-10 关联推荐方法、装置、计算机设备和存储介质

Country Status (6)

Country Link
EP (1) EP3855324A4 (zh)
JP (1) JP7073576B2 (zh)
KR (1) KR20210024173A (zh)
CN (1) CN109446407A (zh)
SG (1) SG11202101614VA (zh)
WO (1) WO2020057022A1 (zh)

Cited By (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111581492A (zh) * 2020-04-01 2020-08-25 车智互联(北京)科技有限公司 一种内容推荐方法、计算设备及可读存储介质
CN111966834A (zh) * 2020-07-29 2020-11-20 深圳市元征科技股份有限公司 一种文件生成方法、文件生成装置及服务器
CN112749344A (zh) * 2021-02-04 2021-05-04 北京百度网讯科技有限公司 信息推荐方法、装置、电子设备、存储介质及程序产品
CN112818258A (zh) * 2021-03-08 2021-05-18 珠海市蜂巢数据技术有限公司 基于关键词的社交网络用户查找方法、计算机装置及计算机可读存储介质
CN112966504A (zh) * 2021-03-26 2021-06-15 深圳壹账通创配科技有限公司 名称识别及关联推荐方法、装置、计算机设备及存储介质
CN113360779A (zh) * 2021-08-09 2021-09-07 智者四海(北京)技术有限公司 内容推荐方法及装置、计算机设备及可读介质
CN113627967A (zh) * 2020-05-09 2021-11-09 北京奇虎科技有限公司 行为数据处理方法、装置、设备及存储介质
CN113656685A (zh) * 2021-07-15 2021-11-16 北京达佳互联信息技术有限公司 一种搜索信息推荐方法、装置、电子设备及存储介质
CN113724817A (zh) * 2021-08-31 2021-11-30 平安国际智慧城市科技股份有限公司 基于人工智能的知识推荐方法、装置、计算机设备及介质
CN113742573A (zh) * 2021-08-03 2021-12-03 深圳Tcl新技术有限公司 一种多媒体资源搜索方法、装置、计算机设备和存储介质
CN113761345A (zh) * 2021-02-09 2021-12-07 北京沃东天骏信息技术有限公司 搜索处理方法、装置、设备及存储介质
CN114065015A (zh) * 2020-07-31 2022-02-18 阿里巴巴集团控股有限公司 搜索推荐方法、装置及设备
CN114138977A (zh) * 2021-12-13 2022-03-04 中国工商银行股份有限公司 日志处理方法、装置、计算机设备和存储介质
CN114356979A (zh) * 2021-11-19 2022-04-15 科大讯飞股份有限公司 一种查询方法及其相关设备
CN114372147A (zh) * 2021-04-07 2022-04-19 北京沃东天骏信息技术有限公司 一种订单日志处理方法和装置
CN115098766A (zh) * 2022-05-26 2022-09-23 安徽省优质采科技发展有限责任公司 一种电子招投标交易平台招投标信息推荐方法及系统
CN115193063A (zh) * 2022-06-27 2022-10-18 广东职业技术学院 一种基于用户偏好的虚拟角色完善方法、设备及介质
CN115794892A (zh) * 2023-01-09 2023-03-14 北京创新乐知网络技术有限公司 基于分层缓存的搜索方法、装置、设备及介质
CN115827986A (zh) * 2022-12-29 2023-03-21 企知道网络技术有限公司 一种技术需求匹配方法、装置、电子设备及存储介质
CN116521975A (zh) * 2022-01-20 2023-08-01 腾讯科技(深圳)有限公司 内容推荐方法、装置、计算机设备、可读存储介质和产品
CN116561438A (zh) * 2023-07-11 2023-08-08 腾讯科技(深圳)有限公司 一种搜索方法和相关装置
CN117093782A (zh) * 2023-10-20 2023-11-21 国网智能科技股份有限公司 一种电力人工智能模型系统及方法
CN117150143A (zh) * 2023-10-30 2023-12-01 华能信息技术有限公司 一种基于工业互联网平台的服务方法及系统
CN117194787A (zh) * 2023-09-11 2023-12-08 中国人民解放军军事科学院战争研究院 一种基于用户行为分析的应用推送方法
CN117194801A (zh) * 2023-11-06 2023-12-08 枣庄市云端互联网产业园管理有限公司 基于技术转移公共服务系统及方法

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110222265B (zh) * 2019-05-28 2022-02-08 深圳市轱辘车联数据技术有限公司 一种信息推送的方法、系统、用户终端及服务器
CN110430477A (zh) * 2019-08-12 2019-11-08 广州华多网络科技有限公司 直播内容的推荐方法、装置、计算机设备和存储介质
CN112632246A (zh) * 2020-12-21 2021-04-09 平安普惠企业管理有限公司 基于深度学习的机器人对话方法、装置和计算机设备
CN116304277B (zh) * 2023-03-01 2023-12-15 张素愿 一种基于ai的智能匹配方法、系统和存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1475936A (zh) * 2002-08-14 2004-02-18 英业达股份有限公司 应用于分布式营业据点的网络社群设立方法以及系统
CN101556588A (zh) * 2008-04-11 2009-10-14 华为技术有限公司 一种搜索方法、设备及系统
CN103916436A (zh) * 2013-01-05 2014-07-09 腾讯科技(深圳)有限公司 信息推送方法、装置、终端及服务器
US20140222870A1 (en) * 2013-02-06 2014-08-07 Lei Zhang System, Method, Software, and Data Structure for Key-Value Mapping and Keys Sorting
CN106407267A (zh) * 2016-08-26 2017-02-15 广州慧睿思通信息科技有限公司 一种基于全文检索的数据分类与数据检索的方法及装置

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009098791A (ja) 2007-10-15 2009-05-07 Canon Inc 文書管理装置
CN101847146A (zh) * 2009-03-27 2010-09-29 华为技术有限公司 一种搜索方法、系统及搜索服务器
JP2015106354A (ja) 2013-12-02 2015-06-08 古河インフォメーション・テクノロジー株式会社 検索サジェスト装置、検索サジェスト方法、及び、プログラム
CN104750718B (zh) * 2013-12-29 2018-06-12 中国移动通信集团公司 一种数据信息的搜索方法和设备
CN108197211A (zh) * 2017-12-28 2018-06-22 百度在线网络技术(北京)有限公司 一种信息推荐方法、装置、服务器和存储介质
CN108227955A (zh) * 2017-12-29 2018-06-29 北京奇虎科技有限公司 一种基于用户历史搜索推荐输入联想的方法及装置
CN108197243A (zh) 2017-12-29 2018-06-22 北京奇虎科技有限公司 一种基于用户身份的输入联想推荐方法及装置
CN108170293A (zh) * 2017-12-29 2018-06-15 北京奇虎科技有限公司 输入联想的个性化推荐方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1475936A (zh) * 2002-08-14 2004-02-18 英业达股份有限公司 应用于分布式营业据点的网络社群设立方法以及系统
CN101556588A (zh) * 2008-04-11 2009-10-14 华为技术有限公司 一种搜索方法、设备及系统
CN103916436A (zh) * 2013-01-05 2014-07-09 腾讯科技(深圳)有限公司 信息推送方法、装置、终端及服务器
US20140222870A1 (en) * 2013-02-06 2014-08-07 Lei Zhang System, Method, Software, and Data Structure for Key-Value Mapping and Keys Sorting
CN106407267A (zh) * 2016-08-26 2017-02-15 广州慧睿思通信息科技有限公司 一种基于全文检索的数据分类与数据检索的方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3855324A4 *

Cited By (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111581492A (zh) * 2020-04-01 2020-08-25 车智互联(北京)科技有限公司 一种内容推荐方法、计算设备及可读存储介质
CN111581492B (zh) * 2020-04-01 2024-02-23 车智互联(北京)科技有限公司 一种内容推荐方法、计算设备及可读存储介质
CN113627967A (zh) * 2020-05-09 2021-11-09 北京奇虎科技有限公司 行为数据处理方法、装置、设备及存储介质
CN111966834A (zh) * 2020-07-29 2020-11-20 深圳市元征科技股份有限公司 一种文件生成方法、文件生成装置及服务器
CN114065015A (zh) * 2020-07-31 2022-02-18 阿里巴巴集团控股有限公司 搜索推荐方法、装置及设备
CN112749344A (zh) * 2021-02-04 2021-05-04 北京百度网讯科技有限公司 信息推荐方法、装置、电子设备、存储介质及程序产品
CN112749344B (zh) * 2021-02-04 2023-08-01 北京百度网讯科技有限公司 信息推荐方法、装置、电子设备、存储介质及程序产品
CN113761345A (zh) * 2021-02-09 2021-12-07 北京沃东天骏信息技术有限公司 搜索处理方法、装置、设备及存储介质
CN112818258B (zh) * 2021-03-08 2024-05-10 珠海市蜂巢数据技术有限公司 基于关键词的社交网络用户查找方法、计算机装置及计算机可读存储介质
CN112818258A (zh) * 2021-03-08 2021-05-18 珠海市蜂巢数据技术有限公司 基于关键词的社交网络用户查找方法、计算机装置及计算机可读存储介质
CN112966504A (zh) * 2021-03-26 2021-06-15 深圳壹账通创配科技有限公司 名称识别及关联推荐方法、装置、计算机设备及存储介质
CN114372147A (zh) * 2021-04-07 2022-04-19 北京沃东天骏信息技术有限公司 一种订单日志处理方法和装置
CN113656685A (zh) * 2021-07-15 2021-11-16 北京达佳互联信息技术有限公司 一种搜索信息推荐方法、装置、电子设备及存储介质
CN113742573A (zh) * 2021-08-03 2021-12-03 深圳Tcl新技术有限公司 一种多媒体资源搜索方法、装置、计算机设备和存储介质
CN113742573B (zh) * 2021-08-03 2023-11-14 深圳Tcl新技术有限公司 一种多媒体资源搜索方法、装置、计算机设备和存储介质
CN113360779A (zh) * 2021-08-09 2021-09-07 智者四海(北京)技术有限公司 内容推荐方法及装置、计算机设备及可读介质
CN113360779B (zh) * 2021-08-09 2023-01-20 智者四海(北京)技术有限公司 内容推荐方法及装置、计算机设备及可读介质
CN113724817A (zh) * 2021-08-31 2021-11-30 平安国际智慧城市科技股份有限公司 基于人工智能的知识推荐方法、装置、计算机设备及介质
CN114356979A (zh) * 2021-11-19 2022-04-15 科大讯飞股份有限公司 一种查询方法及其相关设备
CN114138977A (zh) * 2021-12-13 2022-03-04 中国工商银行股份有限公司 日志处理方法、装置、计算机设备和存储介质
CN116521975A (zh) * 2022-01-20 2023-08-01 腾讯科技(深圳)有限公司 内容推荐方法、装置、计算机设备、可读存储介质和产品
CN115098766A (zh) * 2022-05-26 2022-09-23 安徽省优质采科技发展有限责任公司 一种电子招投标交易平台招投标信息推荐方法及系统
CN115193063A (zh) * 2022-06-27 2022-10-18 广东职业技术学院 一种基于用户偏好的虚拟角色完善方法、设备及介质
CN115827986A (zh) * 2022-12-29 2023-03-21 企知道网络技术有限公司 一种技术需求匹配方法、装置、电子设备及存储介质
CN115794892B (zh) * 2023-01-09 2023-05-23 北京创新乐知网络技术有限公司 基于分层缓存的搜索方法、装置、设备及介质
CN115794892A (zh) * 2023-01-09 2023-03-14 北京创新乐知网络技术有限公司 基于分层缓存的搜索方法、装置、设备及介质
CN116561438A (zh) * 2023-07-11 2023-08-08 腾讯科技(深圳)有限公司 一种搜索方法和相关装置
CN116561438B (zh) * 2023-07-11 2024-01-09 腾讯科技(深圳)有限公司 一种搜索方法和相关装置
CN117194787A (zh) * 2023-09-11 2023-12-08 中国人民解放军军事科学院战争研究院 一种基于用户行为分析的应用推送方法
CN117093782A (zh) * 2023-10-20 2023-11-21 国网智能科技股份有限公司 一种电力人工智能模型系统及方法
CN117093782B (zh) * 2023-10-20 2024-03-12 国网智能科技股份有限公司 一种电力人工智能模型系统及方法
CN117150143A (zh) * 2023-10-30 2023-12-01 华能信息技术有限公司 一种基于工业互联网平台的服务方法及系统
CN117150143B (zh) * 2023-10-30 2024-01-26 华能信息技术有限公司 一种基于工业互联网平台的服务方法及系统
CN117194801B (zh) * 2023-11-06 2024-03-08 枣庄市云端互联网产业园管理有限公司 基于技术转移公共服务系统及方法
CN117194801A (zh) * 2023-11-06 2023-12-08 枣庄市云端互联网产业园管理有限公司 基于技术转移公共服务系统及方法

Also Published As

Publication number Publication date
EP3855324A1 (en) 2021-07-28
JP2021531591A (ja) 2021-11-18
SG11202101614VA (en) 2021-03-30
JP7073576B2 (ja) 2022-05-23
EP3855324A4 (en) 2021-11-03
KR20210024173A (ko) 2021-03-04
CN109446407A (zh) 2019-03-08

Similar Documents

Publication Publication Date Title
WO2020057022A1 (zh) 关联推荐方法、装置、计算机设备和存储介质
US20250362980A1 (en) Systems and methods for censoring text inline
CN110929125B (zh) 搜索召回方法、装置、设备及其存储介质
US10922367B2 (en) Method and system for providing real time search preview personalization in data management systems
US11269665B1 (en) Method and system for user experience personalization in data management systems using machine learning
JP2025526284A (ja) 未構造化文書の管理された要約化および構造化
WO2020077896A1 (zh) 提问数据生成方法、装置、计算机设备和存储介质
CN108664595B (zh) 领域知识库构建方法、装置、计算机设备和存储介质
WO2019136993A1 (zh) 文本相似度计算方法、装置、计算机设备和存储介质
CN112148702B (zh) 一种文件检索的方法及设备
CN112651236B (zh) 提取文本信息的方法、装置、计算机设备和存储介质
CN108595695A (zh) 数据处理方法、装置、计算机设备和存储介质
WO2020206910A1 (zh) 产品信息推送方法、装置、计算机设备和存储介质
WO2021115474A1 (zh) 数据搜索方法、装置、计算机设备和存储介质
CN114399396A (zh) 保险产品推荐方法、装置、计算机设备及存储介质
US20210089667A1 (en) System and method for implementing attribute classification for pii data
CN110362798B (zh) 裁决信息检索分析方法、装置、计算机设备和存储介质
CN110597951A (zh) 文本解析方法、装置、计算机设备和存储介质
CN112559526A (zh) 数据表导出方法、装置、计算机设备及存储介质
CN111753199B (zh) 用户画像构建方法及设备、电子设备和介质
CN113535919A (zh) 一种数据查询的方法、装置、计算机设备以及存储介质
CN113821608B (zh) 服务搜索方法、装置和计算机设备以及存储介质
CN115374849A (zh) 企业相关专利检索方法、装置、设备及介质
CN112732926A (zh) 文本检索方法、装置、计算机设备和存储介质
WO2020057023A1 (zh) 自然语言的语义解析方法、装置、计算机设备和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19863131

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021504196

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 20217003560

Country of ref document: KR

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019863131

Country of ref document: EP

Effective date: 20210419