WO2018113468A1 - Procédé de recommandation de termes de recherche, dispositif, programme et support - Google Patents

Procédé de recommandation de termes de recherche, dispositif, programme et support Download PDF

Info

Publication number
WO2018113468A1
WO2018113468A1 PCT/CN2017/112215 CN2017112215W WO2018113468A1 WO 2018113468 A1 WO2018113468 A1 WO 2018113468A1 CN 2017112215 W CN2017112215 W CN 2017112215W WO 2018113468 A1 WO2018113468 A1 WO 2018113468A1
Authority
WO
WIPO (PCT)
Prior art keywords
search
search term
sensitive
words
user
Prior art date
Application number
PCT/CN2017/112215
Other languages
English (en)
Chinese (zh)
Inventor
彭睿棋
Original Assignee
北京奇虎科技有限公司
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 北京奇虎科技有限公司 filed Critical 北京奇虎科技有限公司
Publication of WO2018113468A1 publication Critical patent/WO2018113468A1/fr

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR 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

Definitions

  • the present invention relates to the field of computer technologies, and in particular, to a search term recommendation method, a search term recommendation device, a computer program, and a computer readable medium.
  • a personalized service area is often set below the search box, for example, a search term frequently input by the user is inserted below the search box.
  • a search term frequently input by the user is inserted below the search box.
  • the user may choose to set the key to turn off the personalized service and no longer display the recommendation. Words, which make the navigation homepage's closing rate higher, resulting in a lower search volume for the site.
  • the solution for setting the personalized service area for the above user is mainly filtered according to the fixed vocabulary of the search engine, and all the searched words including the fixed rule are not displayed, thereby realizing the recommendation of the search word.
  • the recommendation method of the above search words can filter out some sensitive words in the search words, the sensitive words are not displayed in the personalized server area. However, most sensitive words are not common and cannot be filtered by fixed vocabulary rules, making the recommended search terms less accurate. If too many filtering rules are added, a large number of normal search words will be filtered out. , the coverage of recommended search terms cannot be guaranteed, and the amount of recommended search terms is reduced.
  • an embodiment of the present invention provides a search term recommendation method, apparatus, computer program, and computer readable medium, which can improve the accuracy of search term recommendation, thereby improving the amount of recommended search words.
  • the present invention mainly provides the following technical solutions:
  • an embodiment of the present invention provides a search term recommendation method, where the method includes:
  • the recommended search term When receiving the recommended search term sent by the server, the recommended search term is filtered according to the sensitive search term, and the filtered recommended search term is displayed.
  • an embodiment of the present invention further provides a search term recommendation device, where the device includes:
  • An information obtaining module configured to acquire a first search term corresponding to a preset behavior of the user
  • a screening module configured to filter out sensitive search words from the first search words corresponding to the preset behavior
  • the filtering module is configured to: when receiving the recommended search term sent by the server, filter the recommended search term according to the sensitive search term, and display the filtered recommended search term.
  • a computer program comprising computer readable code, when the readable code is run on a computing device, causes the computing device to perform any of the embodiments of the present invention
  • the embodiment of the invention further provides a computer readable medium, wherein the program according to the embodiment of the invention is stored.
  • a search term recommendation method and apparatus provided by an embodiment of the present invention first obtains a first search term corresponding to a preset behavior of a user, where the first search term combines behavior data of a user to search using a search engine, and discovers possible Disturbing the user's search term, and then filtering out the sensitive search words from the first search words corresponding to the preset behavior, further filtering the search words that may involve sensitive information in the first search word as sensitive search words, and receiving the sent by the server When recommending search terms, filter the recommended search terms based on sensitive search terms and display the filtered recommended search terms.
  • the embodiment of the present invention Compared with the prior art method of filtering the vocabulary fixed by the search engine to perform the recommended search term, the embodiment of the present invention reasonably utilizes the first search term corresponding to the user preset behavior, and the user does not want the recommended search. Words are filtered to improve the accuracy of displaying recommended search terms. By filtering the recommended search words based on sensitive search terms, a large number of normal search words are not filtered out, thereby ensuring the recommended amount of search terms from the first A sensitive search term that is filtered out in a search term.
  • FIG. 1 is a flowchart of a search term recommendation method according to an embodiment of the present invention
  • FIG. 2 is a block diagram showing a search term provided by an embodiment of the present invention.
  • FIG. 3 is a flowchart of another search term recommendation method according to an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of a search term recommendation device according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of another search term recommendation apparatus according to an embodiment of the present invention.
  • Figure 6 shows a block diagram of a computing device for performing a search term recommendation method in accordance with the present invention
  • Fig. 7 shows a storage unit for holding or carrying program code implementing the search term recommendation method according to the present invention.
  • An embodiment of the present invention provides a search term recommendation method. As shown in FIG. 1 , the method includes:
  • Step 101 Acquire a first search term corresponding to a preset behavior of the user.
  • the recommended search search form may be a list of search words convenient for the user to query, and the user has visited.
  • the website, the vocabulary selected by a series of recommendation models such as collaborative filtering, etc., the specific recommended search words are shown in Figure 2.
  • the first search term corresponding to the user preset behavior is usually a search term that the user performs a human operation on the undesired recommended search word.
  • the user may be configured to turn off the personalized service and set the recommended search term that is no longer displayed.
  • the deleted search term may be set for the user.
  • the first search term corresponding to the preset behavior is not limited in the embodiment of the present invention, and the first search term corresponding to the preset behavior may be generally represented by the closing rate corresponding to the search term.
  • the closing rate is the ratio of the number of closed users corresponding to the number of displayed users to the search term.
  • the first search term here is often a search term that the user does not expect to display under the search box.
  • the first search term may be related to sensitive information, family privacy, or some users do not like to see.
  • Popular search terms, for example, the vocabulary labeled in Figure 2 may involve sensitive words that the user does not expect to be seen by others, such as ass, underwear, etc., the general user will delete these search terms or set the search recommendation words off. Display, however, this behavioral data may cause some search terms and sensitive vocabulary with good recommendation effects to be deleted or closed at the same time, which will reduce the amount of search words that lead to good recommendation effects, making the recommendation effect unsatisfactory.
  • the user's preference for the recommended search term can be understood, and the search keyword that the user does not want to display is further filtered, thereby improving the user's search experience.
  • Step 102 Filter out sensitive search words from the first search words corresponding to the preset behavior.
  • the first search word corresponding to the preset behavior may include multiple repeated search words, and the higher the number of repeated search words indicates that the user does not expect to display below the search box, the lower the number of search words may indicate the user It may be closed due to an erroneous operation.
  • the search term with a higher number of searches is a search term that the user does not expect to appear.
  • the first search term corresponding to the preset behavior with the number of repetitions being higher than the preset threshold may be used as the sensitive search term, for example, the preset threshold is set to 50, and the first search term corresponding to the preset behavior includes Search terms such as "sexy” and "butt” have a repetition number of 20 for the search term “sexy” and 70 for the "butt", so the search term "sexy” is not a sensitive search term for the search term " Ass is listed as a sensitive search term.
  • the embodiment of the present invention can also determine whether the sensitive keyword is a user by combining the behavior data of a large number of users clicking the recommended search term into the page. What is not expected to be displayed, for example, for the search term "butt", although it is divided into sensitive search words by a preset of repeated number settings, when the user clicks on the search word to enter the page with a high amount of behavior data, the search is required. Whether the word is a sensitive search term for further judgment.
  • the search words that the user does not expect to be recommended are further confirmed, and the search words that the user does not expect to be recommended are filtered, and the recommended search is reduced. Words are bothering users.
  • Step 103 When receiving the recommended search term sent by the server, filtering the recommended search term according to the sensitive search term, and displaying the filtered recommended search term.
  • the recommended search term sent by the server is a search term that is usually calculated based on behavior data of the user search, such as a user browsing a webpage, a search term that the user clicks through the search box, or a recommended search term obtained by using a recommendation algorithm.
  • filtering the recommended search words according to the sensitive search words, and displaying the filtered recommended search words may include, but is not limited to, the following implementation manner, by comparing the sensitive search words with the recommended search words,
  • the recommended search terms contain portions of sensitive search terms that are filtered to preserve the filtered recommended search terms and display the filtered recommended search terms below the search box.
  • the search engine usually filters the recommended search words according to the fixed vocabulary, and displays the filtered recommended search words in the search box.
  • the way to filter the recommended search words through the fixed vocabulary is usually Some sensitive search terms will be filtered out, and search terms that some users may click to search may also be filtered out, making the recommended search terms less effective.
  • a search term recommendation method provided by an embodiment of the present invention first obtains a first search term corresponding to a preset action of a user, where the first search term combines behavior data of a user to search using a search engine, and the user may be disturbed to the user. Search words, and then filter out sensitive search words from the first search words corresponding to the preset behavior, and further filter the search words that may involve sensitive information in the first search words as sensitive search words, when receiving the recommendation sent by the server When searching for words, filter the recommended search terms based on sensitive search terms and display the filtered recommended search terms.
  • the embodiment of the present invention Compared with the prior art method of filtering the vocabulary fixed by the search engine to perform the recommended search term, the embodiment of the present invention reasonably utilizes the first search term corresponding to the user preset behavior, and the user does not want the recommended search. Words are filtered to improve the accuracy of displaying recommended search terms.
  • the method of filtering the recommended search words according to the sensitive search words does not cause a large number of normal search words to be filtered out, thereby ensuring the searched words of the recommended search words are filtered out from the first search words.
  • the embodiment of the present invention provides another search term recommendation method. As shown in FIG. 3, the method includes:
  • Step 301 Obtain a user search for a common word according to historical behavior data searched by the user.
  • the historical behavior data of the user search may include a search term generated by the user through the search behavior, a corresponding page URL, corresponding time information, etc., where the historical behavior data of the user search may be recorded in the user behavior log, according to the history of the user search.
  • the behavior data can understand the user history search data, thereby analyzing the user history search data, and using the search words with frequent user search behavior as the common words searched by the user.
  • the geographically associated search term in the historical behavior data is used as a common search word for the user, if the user is found by analyzing the historical behavior data of the user search. Searching for various life knowledge is more frequent, and the search term for life knowledge in the historical behavior data is used as a common search word for the user.
  • the user searches for common words based on a website frequently visited by the user, and is filtered through a series of recommendation models such as collaborative filtering, and the search words displayed as recommended search words below the search box can be facilitated to some extent.
  • a series of recommendation models such as collaborative filtering
  • Step 302 Acquire a first search term corresponding to the preset behavior of the user according to historical behavior data of the user searching for a common word.
  • the user can search for common words for the user who does not display, and the user does not expect to be displayed, and further obtains the historical behavior data of the user's commonly used search words, such as setting the common search words for the user as The operation of not displaying, setting the deletion operation for the common search words of the user, or setting the hidden operation for the common search words of the user, and further obtaining the first search word corresponding to the preset behavior of the user according to the historical behavior data of the search word by the user, here
  • the default behavior of the preset behavior is not limited by the embodiment of the present invention.
  • a user sets a close search keyword for all users when searching through a search engine, the user is not very satisfied with the commonly used search words displayed, or the user performs through a search engine.
  • you set a search to delete a common search term the user is not very satisfied with the individual search terms displayed.
  • Step 303 Record a repetition rate of the first search term corresponding to the preset behavior.
  • the first search term corresponding to the preset behavior may include multiple identical search terms
  • the same search term indicates that the user has a poor preference for the search term in the process of performing the search, and may be inferred that the user does not expect to display.
  • the search term if the repetition rate is higher, indicates that the user's preference for the search term is worse, and the repetition rate of each search word in the first search term is further recorded.
  • Step 304 The second search term whose repetition rate is higher than the preset threshold in the first search word corresponding to the preset behavior is used as the sensitive search term.
  • the search words in the first search word are further filtered, and the filtered user may be divided into the first search word due to an erroneous operation.
  • the search term is not clicked or even expected to be displayed during the search process. It should be noted that the preset threshold is not in the embodiment of the present invention. It can be set according to actual needs.
  • Step 305 Compare the second search term with a popular search term, and use the second search term that is specific to the sensitive search term.
  • the first search data of the user's preset behavior may include recent popular search terms, such as the hit dramas that most users have recently searched for: Jinxiu Weiyang, ghost Blowing Light, etc.
  • the second search term is further compared with the popular search term, so that the second search term coincident with the comparison is used as the sensitive search term.
  • the second search term combined with the specific search term is used as a sensitive search term, and the popular search term of the user preset behavior is reasonably utilized to further judge the user's search term.
  • the degree of preference avoids the influence of recommending sensitive search terms on users, so that users can obtain satisfactory search results by recommending search terms.
  • Step 306 When receiving the recommended search term sent by the server, filtering the recommended search term according to the sensitive search term, and displaying the filtered recommended search term.
  • the recommended search words are filtered according to the sensitive search words, and the filtered recommended search words are displayed.
  • the recommended search words may be included in the recommended search words, for example, the recommended search words are sexy legs, and sensitive.
  • the search term includes sexy, then judge the recommended search
  • the word contains sensitive search words, and further filters the sensitive search words in the recommended search words, and displays the filtered recommended search words, such as the filtered recommended search words are beautiful legs, if the recommended search words do not contain sensitive search Words are not filtered and show recommended search terms.
  • Step 307 Calculate sensitivity of the sensitive search term according to a repetition rate of the sensitive search term.
  • the sensitivity of the sensitive search words can be calculated according to the repetition rate of the sensitive search words, and the sensitivity can be used as a basis for evaluating sensitive search words. For sensitive search words with low sensitivity, as the number of searched users increases, when the number of sensitive search words increases, it may be considered to exclude sensitive search words as a general search word.
  • Step 308 Sort the sensitive search words according to sensitivity, and establish a sensitive vocabulary according to the sorted sensitive search words.
  • the sensitive search words recorded in the sensitive lexicon are mostly search words that the user does not expect to see, and the user closes the search words in the frequently searched words, and records the sensitivity of each user's frequently searched words, thereby
  • the sensitivity ranking establishes sensitive lexicons. For example, the sensitivity rankings of sensitive search terms in sensitive lexicon are ranked from high to low as shown in Table 1 below.
  • the recommended search term is further filtered by establishing a sensitive vocabulary, and the recommended search term includes the sensitive vocabulary.
  • Sensitive search words directly filter the parts containing sensitive search terms to facilitate the processing of search recommendation words.
  • Step 309 Update the sensitive search term in the sensitive thesaurus according to a preset time interval.
  • updating the sensitive search term in the sensitive vocabulary according to the preset time interval may include, but is not limited to, the following implementation manner, and the sensitive vocabulary is updated according to the sensitivity of the sensitive search term in the sensitive vocabulary.
  • the user's closed search term may contain recent popular search terms, and may also be combined with popular search terms to increase sensitive search terms in the sensitive lexicon, such as a recent popular microblog or some Popular movies, you can confirm whether to add the popular search words to the sensitive lexicon through a large number of users' recent searches.
  • the specific application scenario may be as follows, but is not limited thereto, including: firstly, filtering the top ranked search words according to the number of user history search words, and recommending the top ranked search words through recommendation.
  • the model filters out some illegal sensitive vocabulary, it retains a number of search terms as user frequent search words to display under the search engine search box. In order to confirm whether the user often matches the user's expectations, it is further closed according to the search term.
  • the historical behavior data of the operation acquires a first search term corresponding to the closing operation
  • the first search term is usually a search term that the user does not expect to display, and then in order to continue to filter from the first search term, the first corresponding to the closing operation is recorded
  • the repetition rate of the search word is used as a sensitive search term for the second search term with a repetition rate higher than 30%.
  • the current popular search may also be combined. Add search terms from popular searches to sensitive search terms, where sensitive search terms are based
  • the revival value can be ranked, and the lexicon or vocabulary can be built according to the sensitive search words to facilitate viewing.
  • the sensitive vocabulary or the sensitive search words in the vocabulary can be updated according to the sensitivity of the sensitive search words.
  • Sensitive search words with low sensitivity in sensitive lexicon are deleted, sensitive search words with higher sensitivity are retained, and search words with higher closing rate in popular search are added to sensitive vocabulary, and finally based on sensitive search words
  • the search words are compared, and the sensitive search words included in the frequently searched words are filtered, so that when the user clicks on the 360 search engine to search, the filtered frequently searched words are displayed below the search box.
  • Another search term recommendation method in the embodiment of the present invention can reasonably utilize search words that are closed or deleted by the user during the search process, thereby mining search words that may disturb users or involve personal privacy and sensitive information, and further sensitive search terms. Filtering, so that the filtered search terms are displayed to the user as recommended search words, can improve the user's click rate on the search words, and let the user search for satisfactory results.
  • the embodiment provides an apparatus embodiment corresponding to the foregoing method embodiment.
  • the device shows a search term recommendation device, and the device may include:
  • the information obtaining module 41 is configured to obtain a first search term corresponding to the preset behavior of the user, where the information acquiring module 41 is a main function module for acquiring the first search term corresponding to the preset behavior of the user in the search term recommendation device.
  • the first search term corresponding to the preset behavior may be generally represented by a closing rate corresponding to the search term, where the first search term may be some popular information related to sensitive information, family privacy, or some users do not like to see. Search word
  • the screening module 42 is configured to filter the sensitive search term from the first search term corresponding to the preset behavior, and the screening module 42 is a main functional module for filtering the sensitive search term in the search term recommendation device.
  • the first search term corresponding to the preset behavior with the number of repetitions being higher than the preset threshold is used as the sensitive search term, and the search term corresponding to the behavior data of the user entering the page by clicking the recommended search term may be further combined as the sensitive search term;
  • the filtering module 43 is configured to: when receiving the recommended search term sent by the server, filter the recommended search term according to the sensitive search term, and display the filtered recommended search term, where the filtering module 43 is a search
  • the main function module for filtering the recommended search words in the word recommendation device specifically by comparing the sensitive search words with the recommended search words, filtering the parts of the recommended search words containing the sensitive search words, thereby retaining the filtered Recommended search terms.
  • a search term recommendation device provided by an embodiment of the present invention first obtains a first search term corresponding to a preset behavior of a user, where the first search term combines the behavior data of the user to search using the search engine, and the user may be disturbed to the user. Search words, and then filter out sensitive search words from the first search words corresponding to the preset behavior, and further filter the search words that may involve sensitive information in the first search words as sensitive search words, when receiving the recommendation sent by the server When searching for words, filter the recommended search terms based on sensitive search terms and display the filtered recommended search terms.
  • the first search term corresponding to the preset action of the user is used reasonably, and the search term that the user does not want to be recommended is compared with the method of performing the search by the search engine fixed vocabulary. Filtering, so as to improve the accuracy of displaying recommended search words, by filtering the recommended search words according to sensitive search words, does not cause a large number of normal search words to be filtered out, thereby ensuring the recommended amount of recommended search words from the first Sensitive search terms filtered out in search terms.
  • an embodiment of the present invention provides another search term recommendation device, where the device further includes:
  • the comparison module 44 can be configured to compare the second search term with a popular search term, and use the second search term that is specific to the sensitive search term, and the comparison module 44 is used in a search term recommendation device. According to the comparison between the second search word and the popular search word as the main function module of the sensitive search word, the popular search term of the user preset behavior is used reasonably to further judge the user's preference for the search word;
  • the calculating module 45 is configured to calculate a sensitivity of the sensitive search term according to a repetition rate of the sensitive search term, and the calculating module 45 is a main functional module for calculating a sensitivity of the sensitive search term in the search term recommendation device. ;
  • the establishing module 46 is configured to sort the sensitive search words according to the sensitivity, and establish a sensitive vocabulary according to the sorted sensitive search words.
  • the establishing module 46 is used in a search term recommendation device to establish a sensitive vocabulary.
  • the main function module specifically places the filtered sensitive search words in a sensitive vocabulary or medium to facilitate subsequent searching;
  • the update module 47 is configured to update the sensitive search term in the sensitive vocabulary according to the preset time interval, and the update module 47 is a main function module for updating the sensitive vocabulary in the search term recommendation device, which may be specifically Sensitivity of sensitive search terms in sensitive thesaurus updates sensitive thesaurus, and can also be combined with popular search terms to increase sensitive search terms in sensitive thesaurus.
  • the information acquiring module 41 includes:
  • the first obtaining module 411 is configured to obtain a user search for a common word according to the historical behavior data searched by the user;
  • the second obtaining module 412 is configured to obtain a first search term corresponding to the preset behavior of the user according to historical behavior data of the user searching for a common word by the user.
  • the screening module 42 includes:
  • the recording module 421 can be configured to record a repetition rate of the first search term corresponding to the preset behavior
  • the screening module 422 is configured to use, as the sensitive search term, a second search time in which the repetition rate of the first search word corresponding to the preset behavior is higher than a preset threshold.
  • the filtering module 43 includes:
  • the determining module 431 can be configured to determine whether the sensitive search term is included in the recommended search term
  • the first display module 432 may be configured to: if the sensitive search term is included in the recommended search term, filter the sensitive search term in the recommended search term to display the filtered recommended search term;
  • the second display module 433 can be configured to display the search keyword if the sensitive search term is not included in the recommended search term.
  • the updating module 47 is specifically configured to delete, according to the preset time interval, the sensitive search term in the sensitive vocabulary with a lower sensitivity ranking.
  • Another search term recommendation device of the embodiment of the present invention can reasonably utilize search words that are closed or deleted by the user during the search process, thereby searching for search words that may disturb users or involve personal privacy and sensitive information, and further sensitive search terms. Filtering, so that the filtered search terms are displayed to the user as recommended search words, can improve the user's click rate on the search words, and let the user search for satisfactory results.
  • the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment.
  • modules in the devices of the embodiments can be adaptively changed and placed in one or more devices different from the embodiment.
  • the modules or units or components of the embodiments may be combined into one module or unit or component, and further they may be divided into a plurality of sub-modules or sub-units or sub-components.
  • any combination of the features disclosed in the specification, including the accompanying claims, the abstract and the drawings, and any methods so disclosed, or All processes or units of the device are combined.
  • Each feature disclosed in this specification (including the accompanying claims, the abstract and the drawings) may be replaced by alternative features that provide the same, equivalent or similar purpose.
  • the various component embodiments of the present invention may be implemented in hardware, or in a software module running on one or more processors, or in a combination thereof.
  • a microprocessor or digital signal processor may be used in practice to implement some or all of some or all of the methods and apparatus for data storage in accordance with embodiments of the present invention.
  • the invention can also be implemented as a device or device program (e.g., a computer program and a computer program product) for performing some or all of the methods described herein.
  • Such a program implementing the invention may be stored on a computer readable medium or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.
  • Figure 6 illustrates a computing device that can implement a search term recommendation method in accordance with the present invention.
  • the computing device conventionally includes a processor 610 and a program product or readable medium in the form of a memory 620.
  • Memory 620 can be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, or ROM.
  • Memory 620 has a memory space 630 for program code 631 for performing any of the method steps described above.
  • storage space 630 for program code may include various program code 631 for implementing various steps in the above methods, respectively.
  • These program codes can be read from or written to one or more program products.
  • These program products include program code carriers such as memory cards.
  • Such a program product is typically a portable or fixed storage unit as described with reference to FIG.
  • the storage unit may have storage segments, storage spaces, and the like that are similarly arranged to memory 620 in the computing device of FIG.
  • the program code can be compressed, for example, in an appropriate form.
  • the storage unit includes readable code 631', ie, code that can be read by a processor, such as 610, which, when executed by a computing device, causes the computing device to perform various steps in the methods described above. .

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

L'invention concerne un procédé de recommandation de termes de recherche, un dispositif, un programme et un support se rapportant au domaine de la technologie informatique et étant utilisés pour améliorer la précision de la recommandation de termes de recherche afin d'augmenter la quantité d'exportation du terme de recherche recommandé. Le procédé consiste : à obtenir un premier terme de recherche correspondant à un comportement d'utilisateur prédéfini ; à filtrer un terme de recherche sensible à partir du premier terme de recherche correspondant au comportement d'utilisateur prédéfini ; et lorsqu'un terme de recherche recommandé envoyé par un serveur est reçu, à filtrer le terme de recherche recommandé en fonction du terme de recherche sensible et à afficher le terme de recherche recommandé filtré. La présente invention est principalement utilisée pour la recommandation de termes de recherche.
PCT/CN2017/112215 2016-12-23 2017-11-22 Procédé de recommandation de termes de recherche, dispositif, programme et support WO2018113468A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611213479.2A CN106777217B (zh) 2016-12-23 2016-12-23 一种搜索词推荐方法及装置
CN201611213479.2 2016-12-23

Publications (1)

Publication Number Publication Date
WO2018113468A1 true WO2018113468A1 (fr) 2018-06-28

Family

ID=58920876

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/112215 WO2018113468A1 (fr) 2016-12-23 2017-11-22 Procédé de recommandation de termes de recherche, dispositif, programme et support

Country Status (2)

Country Link
CN (1) CN106777217B (fr)
WO (1) WO2018113468A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109858344A (zh) * 2018-12-24 2019-06-07 深圳市珍爱捷云信息技术有限公司 婚恋对象推荐方法、装置、计算机设备及存储介质
CN111080339A (zh) * 2019-11-18 2020-04-28 口口相传(北京)网络技术有限公司 基于场景的类目偏好数据生成方法及装置
CN111611491A (zh) * 2020-05-25 2020-09-01 深圳前海微众银行股份有限公司 搜索词推荐方法、装置、设备及可读存储介质

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106777217B (zh) * 2016-12-23 2020-10-30 北京奇虎科技有限公司 一种搜索词推荐方法及装置
CN107220306B (zh) * 2017-05-10 2021-09-28 百度在线网络技术(北京)有限公司 一种搜索方法和装置
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
CN109189990B (zh) * 2018-07-25 2021-03-26 北京奇艺世纪科技有限公司 一种搜索词的生成方法、装置及电子设备
CN111723280B (zh) * 2019-03-20 2023-06-16 北京字节跳动网络技术有限公司 信息的处理方法、装置、存储介质及电子设备
CN113761170A (zh) * 2020-09-15 2021-12-07 北京沃东天骏信息技术有限公司 更新语料库的方法和装置
CN115935069A (zh) * 2022-12-29 2023-04-07 北京字跳网络技术有限公司 一种搜索词确定方法、装置、计算机设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103294814A (zh) * 2013-06-07 2013-09-11 百度在线网络技术(北京)有限公司 搜索结果推荐方法、系统和搜索引擎
US20140379745A1 (en) * 2010-12-31 2014-12-25 Alibaba Group Holding Limited Recommendation of search keywords based on indication of user intention
CN104933081A (zh) * 2014-03-21 2015-09-23 阿里巴巴集团控股有限公司 一种搜索建议提供方法及装置
CN105069168A (zh) * 2015-08-28 2015-11-18 百度在线网络技术(北京)有限公司 搜索词推荐方法和装置
CN105956149A (zh) * 2016-05-12 2016-09-21 北京奇艺世纪科技有限公司 默认搜索词的推荐方法和装置
CN106777217A (zh) * 2016-12-23 2017-05-31 北京奇虎科技有限公司 一种搜索词推荐方法及装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104572889B (zh) * 2014-12-24 2016-10-05 深圳市腾讯计算机系统有限公司 一种搜索词推荐方法、装置和系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140379745A1 (en) * 2010-12-31 2014-12-25 Alibaba Group Holding Limited Recommendation of search keywords based on indication of user intention
CN103294814A (zh) * 2013-06-07 2013-09-11 百度在线网络技术(北京)有限公司 搜索结果推荐方法、系统和搜索引擎
CN104933081A (zh) * 2014-03-21 2015-09-23 阿里巴巴集团控股有限公司 一种搜索建议提供方法及装置
CN105069168A (zh) * 2015-08-28 2015-11-18 百度在线网络技术(北京)有限公司 搜索词推荐方法和装置
CN105956149A (zh) * 2016-05-12 2016-09-21 北京奇艺世纪科技有限公司 默认搜索词的推荐方法和装置
CN106777217A (zh) * 2016-12-23 2017-05-31 北京奇虎科技有限公司 一种搜索词推荐方法及装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109858344A (zh) * 2018-12-24 2019-06-07 深圳市珍爱捷云信息技术有限公司 婚恋对象推荐方法、装置、计算机设备及存储介质
CN111080339A (zh) * 2019-11-18 2020-04-28 口口相传(北京)网络技术有限公司 基于场景的类目偏好数据生成方法及装置
CN111080339B (zh) * 2019-11-18 2024-01-30 口口相传(北京)网络技术有限公司 基于场景的类目偏好数据生成方法及装置
CN111611491A (zh) * 2020-05-25 2020-09-01 深圳前海微众银行股份有限公司 搜索词推荐方法、装置、设备及可读存储介质

Also Published As

Publication number Publication date
CN106777217A (zh) 2017-05-31
CN106777217B (zh) 2020-10-30

Similar Documents

Publication Publication Date Title
WO2018113468A1 (fr) Procédé de recommandation de termes de recherche, dispositif, programme et support
US10289618B2 (en) Third party search applications for a search system
JP6622227B2 (ja) ユーザ関係データユーザ関係データの結合に基づく検索
CN107145496B (zh) 基于关键词将图像与内容项目匹配的方法
CN107480158B (zh) 基于相似性得分评估内容项目与图像的匹配的方法和系统
CN107463591B (zh) 响应于搜索查询对待与内容匹配的图像动态排序的方法和系统
JP5575902B2 (ja) クエリのセマンティックパターンに基づく情報検索
US9720913B1 (en) Automatic generation of snippets based on context and user interest
US8849812B1 (en) Generating content for topics based on user demand
US10452728B2 (en) Method and apparatus for recommending keywords
CN107766399B (zh) 用于使图像与内容项目匹配的方法和系统及机器可读介质
US11693910B2 (en) Personalized search result rankings
US20190155846A1 (en) Search result displaying method and apparatus
US20160283952A1 (en) Ranking information providers
WO2016187705A1 (fr) Système et procédé de classement de résultats de recherche
US20150347543A1 (en) Federated search
US20150347503A1 (en) Multi-domain query completion
US20190163809A1 (en) Streaming events analysis for search recall improvements
WO2014164665A2 (fr) Intégration d'offres
CN106874335A (zh) 行为数据处理方法、装置及服务器
EP2933734A1 (fr) Procédé et système d'analyse structurelle de sites Web
CN106599299A (zh) 一种网站关键词的确定方法及装置
US20170323019A1 (en) Ranking information providers
WO2015149550A1 (fr) Procédé et appareil pour déterminer des notes de liens dans un site internet
US20150269268A1 (en) Search server and search method

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: 17883808

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17883808

Country of ref document: EP

Kind code of ref document: A1