CN113127736A - Classification recommendation method and device based on search history - Google Patents
Classification recommendation method and device based on search history Download PDFInfo
- Publication number
- CN113127736A CN113127736A CN202110398716.1A CN202110398716A CN113127736A CN 113127736 A CN113127736 A CN 113127736A CN 202110398716 A CN202110398716 A CN 202110398716A CN 113127736 A CN113127736 A CN 113127736A
- Authority
- CN
- China
- Prior art keywords
- page
- specified
- label
- user
- classification
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
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)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a network searching technology, and provides a classified recommendation method, a device, computer equipment and a storage medium based on a search history record, wherein the method comprises the following steps: s1: receiving a specified keyword input by a specified user; s2: traversing the search history records of all users in the database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is obtained according to the specified keyword search and is browsed by the user; s3: acquiring first designated tags corresponding to all clicked pages; s4: matching the appointed keywords with each first appointed label to obtain a second appointed label with the highest matching degree with the appointed keywords; s5: finding a first recommended page from the second specified tag; s6: the first recommendation page is recommended to the designated user, the method not only improves the searching efficiency of the user, but also can ensure that the content displayed to the user is the content which can be clicked by the user as far as possible, and the requirements of the user are met.
Description
Technical Field
The invention relates to the technical field of network search, in particular to a classification recommendation method and device based on search history records, computer equipment and a storage medium.
Background
With the rapid development of scientific technology, the relationship between the internet and people is increasingly tight, the corresponding internet data is also massively and explosively increased, and meanwhile, people can actively or passively acquire required information from the massive information, wherein when the information is actively acquired through a network channel, the information is usually acquired through various website search engines, for example, keywords are input through search engines such as Baidu and Google, and then a large number of recommendation results with high similarity are acquired according to the keywords, but as a result, the search results without pertinence are listed, generally, valuable data are more, so that a user needs to input the keywords for many times or change the keywords or change the character positions of the keywords for searching, and the complexity and the efficiency are low, and the searched information is not necessarily consistent with the expectation, and the user requirements are difficult to meet.
Disclosure of Invention
The invention mainly aims to provide a classification recommendation method, a classification recommendation device, computer equipment and a storage medium based on search history records, and aims to solve the technical problem that invalid information is easy to recommend by searching keywords in the prior art, so that the user requirements are difficult to meet.
Based on the above object, the present invention provides a classification recommendation method based on search history, comprising:
s1: receiving a specified keyword input by a specified user;
s2: traversing the search history records of all users in a database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is searched according to the specified keyword and is browsed by the user;
s3: obtaining classification tags corresponding to all the clicked pages, recording the classification tags as first designated tags, wherein different classification tags correspond to different types of pages, and the same classification tag corresponds to a plurality of different similar pages;
s4: matching the appointed keywords with each first appointed label to obtain a label with the highest matching degree with the appointed keywords, and marking as a second appointed label;
s5: finding a first recommended page from the second specified tag;
s6: and recommending the first recommendation page to the specified user.
Further, the step S5 includes:
s50: and searching the clicked page with the highest click frequency from the second appointed label, and recording the clicked page as the first recommended page.
Further, after the step S4, the method includes:
s40: and obtaining a label with the second highest matching degree with the specified keyword, recording the label as a third specified label, finding the clicked page with the highest clicking frequency from the third specified label, recording the clicked page as a second recommended page, repeating the traversing of all the first specified labels until the clicked page with the highest clicking frequency in each first specified label is found, obtaining a plurality of recommended pages, and recommending the recommended pages to the specified user.
Further, after the step S6, the method includes:
s7: after the classification tags of the currently found recommended pages are removed, matching the specified keywords with the rest of the first specified tags to obtain the classification tags with the highest matching degree with the specified keywords in the rest of the first specified tags;
s8: finding a click page with the highest click frequency from the classification label with the highest matching degree with the specified keyword, and recording the click page as a third recommendation page;
s9: and repeating the steps S7-S8 to obtain a plurality of third recommendation pages, and recommending the third recommendation pages to the specified user.
Further, the step S5 includes:
s51: acquiring a search history of the specified user;
s52: extracting the user characteristics of the specified user from the search history record;
s53: inputting the user characteristics into a preset prediction model for calculation to obtain a prediction result, wherein the prediction model is a neural network model obtained by training by taking the user characteristics and a corresponding page as a training set;
s54: and determining the first recommended page from the second designated label according to the prediction result.
Further, the step S4 includes:
s41: acquiring page information of all clicked pages in the first appointed label;
s42: extracting abstract information from the page information, and generating label information by the abstract information and the label name of the first specified label;
s43: matching the appointed keywords with the label information of each first appointed label to obtain a corresponding matching degree;
s44: and sorting the first appointed labels from high to low according to the matching degree to obtain the second appointed label with the highest matching degree.
The invention also provides a classified recommendation device based on the search history record, which comprises:
a receiving key unit for receiving a specified keyword input by a specified user;
the traversal recording unit is used for traversing the search history records of all users in the database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is searched according to the specified keyword and is browsed by the user;
the system comprises a tag obtaining unit, a tag obtaining unit and a tag selecting unit, wherein the tag obtaining unit is used for obtaining classification tags corresponding to all clicked pages and recording the classification tags as first designated tags, different classification tags correspond to different types of pages, and the same classification tag corresponds to a plurality of different similar pages;
a matching label unit, configured to match the specified keyword with each of the first specified labels, so as to obtain a label with a highest matching degree with the specified keyword, and mark the label as a second specified label;
the page finding unit is used for finding a first recommended page from the second specified label;
and the recommendation page unit is used for recommending the first recommendation page to the specified user.
Further, the finding a page unit includes:
and the click finding subunit is used for finding the click page with the highest click frequency from the second specified label and recording the click page as the first recommended page.
The present invention also provides a computer-readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the above-mentioned search history-based classification recommendation method.
The present invention also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the search history based classification recommendation method of any one of claims 1 to 6.
The invention has the beneficial effects that: when the input appointed keywords are received, traversing all the appointed keywords searched by the user and the classification labels of the click results to find the classification label with the highest matching degree, and recommending the page with the highest frequency in the classification labels to the user, so that the searching efficiency of the user is improved, the content displayed to the user can be ensured to be the content which can be clicked by the user as far as possible, and the requirement of the user is met.
Drawings
FIG. 1 is a diagram illustrating steps of a search history based classification recommendation method according to an embodiment of the present invention;
FIG. 2 is a block diagram illustrating a structure of a classification recommendation device based on search history according to an embodiment of the present invention;
FIG. 3 is a block diagram illustrating the structure of one embodiment of a storage medium of the present application;
FIG. 4 is a block diagram illustrating the structure of one embodiment of a computer device of the present application.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, the method for recommending a category based on a search history in this embodiment includes:
step S1: receiving a specified keyword input by a specified user;
step S2: traversing the search history records of all users in a database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is searched according to the specified keyword and is browsed by the user;
step S3: obtaining classification tags corresponding to all the clicked pages, recording the classification tags as first designated tags, wherein different classification tags correspond to different types of pages, and the same classification tag corresponds to a plurality of different similar pages;
step S4: matching the appointed keywords with each first appointed label to obtain a label with the highest matching degree with the appointed keywords, and marking as a second appointed label;
step S5: finding a first recommended page from the second specified tag;
step S6: and recommending the first recommendation page to the specified user.
The classification recommendation method based on the search history is applied to various websites with search engines, such as Taobao, Baidu, Google and the like. In this embodiment, the specified user refers to a user who currently inputs a keyword in a search engine, and the specified keyword may be a keyword currently input in the search engine, for example, a phrase or a sentence such as "jeans", "congratulations for getting rich", or "white rabbit".
As described in steps S1-S2, after receiving the specified keyword input by the specified user, traversing the search history of all users in the database, finding out the user who has input the specified keyword, and according to the click result of the specified keyword, specifically, the page that is searched according to the specified keyword and browsed by the user, for convenience of description, the page is called the click page, it should be understood that, before receiving the specified keyword input by the specified user, each user inputs the search keyword on the search page, jumps to the search result page, and when clicking on a single result, records the current search keyword of the user and the information of the click page, and stores the current search keyword and the information of the click page in the database.
For example, if the specified user inputs "jeans", the database is traversed by all users to obtain a plurality of different users who have searched for "jeans", and the recommended pages obtained by searching for "jeans" include a "purchase page for purchasing jeans", "teaching page for wearing with jeans matching", "introduction page for introducing jeans brand", and the like.
In step S3, obtaining classification tags corresponding to all clicked pages, in order to facilitate distinguishing the first designated tag, in this embodiment, the classification tags are assigned to all pages, a tag system is established, different classification tags correspond to different types of pages, for example, types including shopping, teaching, introduction, and the like, the same classification tag corresponds to a plurality of different similar pages, for example, "purchase page for buying jeans" and "purchase page for buying congratulatory fortune charm" are different similar pages in the same classification tag "shopping", and each classification tag carries corresponding tag information for matching with a keyword.
Further, the classification labels may be classified into classes, such as a first class classification, a second class classification, a third class classification, a first class classification of "shopping", a second class classification of "purchasing clothes", "purchasing electronic products", a third class classification of "purchasing trousers", "purchasing jacket", and the like.
As shown in step S4, after the first designated tag is found, the designated keyword is matched with the first designated tag, so as to find the tag with the highest matching degree with the designated keyword, specifically, the designated keyword is matched with the tag information of the first designated tag, and the tag can be matched in a combination of multiple matching manners, or one matching manner is directly adopted for matching, and after the tag with the highest matching degree is obtained by matching, the tag is marked as the second designated tag.
In steps S5-S6, the first recommended page is found from the second designated tag, and the recommended page is recommended to the designated user, it should be understood that the second designated tag includes a plurality of clicked pages, and the first recommended page can be found by the click rate or according to the rule set by the user attribute, and then recommended to the designated user. Specifically, in one embodiment, the step S5 includes:
step S50: and searching the clicked page with the highest click frequency from the second appointed label, and recording the clicked page as the first recommended page.
In this embodiment, the second designated tab is associated with a plurality of different pages, which include the above clicked pages, and the clicked page with the highest click frequency, that is, the above first recommended page, can be found from the clicked pages, and it is noted that the click frequency here refers to the click of the user based on the input of the designated keyword search, rather than the click based on the input of other keywords search, for example, the designated keyword is "jeans", and after matching the classification tag of the shopping, the total click of "purchase page of jeans" is 100 times, and in a certain time, the click of "purchase page of jeans" is 60 times based on the input of the designated keyword "jeans", the click of "purchase page of jeans" is 40 times based on the input of the "jeans" keyword, and the total click of "purchase page of jeans" is 150 times, and the click of "is 50 times based on the input of the designated keyword of" jeans ", the click rate is calculated to be the highest and is still the "purchase page for buying jeans" here, even though the total click rate of the "purchase page for buying jeans" is larger than the total click rate of the "purchase page for buying jeans" based on the "shorts" keyword for 100 clicks.
In another embodiment, in order to improve the matching rate with the specified user, finding the first recommended page may be performed based on an attribute of the specified user, and specifically, the step S5 includes:
step S51: acquiring a search history of the specified user;
step S52: extracting the user characteristics of the specified user from the search history record;
step S53: inputting the user characteristics into a preset prediction model for calculation to obtain a prediction result, wherein the prediction model is a neural network model obtained by training by taking the user characteristics and a corresponding page as a training set;
step S54: and determining the first recommended page from the second designated label according to the prediction result.
In the embodiment, a search history of a specified user is obtained, wherein the search history includes search keywords and browsed pages, user characteristics of the specified user are extracted by counting the search keywords, the types of the browsed pages and corresponding search times, the user characteristics are used for representing search preferences of the specified user, the user characteristics are input into a preset prediction model to be calculated to obtain a prediction result, the prediction result is the probability that the specified user may click each page, and the page with the maximum probability is used as a first recommended page. The prediction module is a neural network model formed by deep learning and training through a training set, wherein samples in the training set comprise user characteristics and pages which are most likely to be clicked by a user, and a specific training process is the prior art and is not repeated here.
In another embodiment, other clicked pages in the second designated tag can be ranked from high to low according to the probability, and then the clicked page with the preset ranking rank is selected as a recommended page and recommended to the designated user.
According to the classification recommendation method based on the search history provided by the invention, when the input specified keyword is received, the specified keyword searched by all users and the classification label of the click result are traversed, the classification label with the highest matching degree is found, and the page with the highest frequency in the classification label is recommended to the user, so that the search efficiency of the user is improved, the content displayed to the user can be ensured to be the content which can be clicked by the user as much as possible, and the requirements of the user are met.
In one embodiment, after the step S4, the method further includes:
step S40: and obtaining a label with the second highest matching degree with the specified keyword, recording the label as a third specified label, finding the clicked page with the highest clicking frequency from the third specified label, recording the clicked page as a second recommended page, repeating the traversing of all the first specified labels until the clicked page with the highest clicking frequency in each first specified label is found, obtaining a plurality of recommended pages, and recommending the recommended pages to the specified user.
In this embodiment, in order to further improve the search time of the user, find the content required by the user as much as possible, recommend the page with the highest click frequency in each classification tag as a recommended page to the specified user, specifically, after finding the first recommended page, obtain the tag with the second highest matching degree with the specified keyword again, then find the clicked page with the highest click frequency from the obtained clicked pages, and so on, obtain the tags with the third highest matching degree with the specified keyword respectively, and obtain the recommended page of each first specified tag until all the first specified tags are traversed by the tag with the fourth highest matching degree with the specified keyword, but certainly, the recommended page of each first specified tag may be searched without searching in the order of the matching degrees, and only all the first specified tags need to be traversed.
In another embodiment, after the step S6, the method further includes:
step S7: after the classification tags of the currently found recommended pages are removed, matching the specified keywords with the rest of the first specified tags to obtain the classification tags with the highest matching degree with the specified keywords in the rest of the first specified tags;
step S8: finding a click page with the highest click frequency from the classification label with the highest matching degree with the specified keyword, and recording the click page as a third recommendation page;
step S9: and repeating the steps S7-S8 to obtain a plurality of third recommendation pages, and recommending the third recommendation pages to the specified user.
In this embodiment, for more accurate matching, after a corresponding recommended page is found in each classification tag, the classification tag may be removed, and then re-matching is performed, specifically, after the recommended page is found through the first specified tag, the first specified tag is removed, then, the remaining first specified tags are matched with the specified keywords again, the classification tag with the highest matching degree is obtained again, so that a third recommended page is obtained, the above steps are repeated until all the first specified tags are matched, a plurality of third recommended pages are obtained, sorting may be performed according to the order of obtaining the recommended pages, and then, the recommended pages sorted within the preset ranking are selected and recommended to the user.
In one embodiment, the step S4 includes:
step S41: acquiring page information of all clicked pages in the first appointed label;
step S42: extracting abstract information from the page information, and generating label information by the abstract information and the label name of the first specified label;
step S43: matching the appointed keywords with the label information of each first appointed label to obtain a corresponding matching degree;
step S44: and sorting the first appointed labels from high to low according to the matching degree to obtain a second appointed label with the highest matching degree.
In this embodiment, in the step S4, the second specified label with the highest matching degree is obtained by matching the specified keyword with the first specified label, specifically, the page information of all clicked pages in the first specified label is obtained first, then extracting corresponding features from all page information, assembling all the features into abstract information, and generating label information by the abstract information and the label name of the first appointed label, of course, in order to further improve matching precision, information such as label attribute can be added, then matching the appointed keywords with the tag information to obtain the matching degree of each first appointed tag and the appointed keywords, then sequencing according to the matching degree from high to ground to obtain a second appointed tag with the highest matching degree, therefore, the matching precision of the keywords and the page is further improved by adding the information related to the page into the tag information.
Referring to fig. 2, in this embodiment, a search history based classification recommendation apparatus corresponding to the search history based classification recommendation method is provided, and the apparatus includes:
a receiving key unit 1 for receiving a specified keyword input by a specified user;
the traversal recording unit 2 is configured to traverse search history records of all users in the database to obtain a clicked page corresponding to the specified keyword, where the clicked page is a page that is obtained by searching according to the specified keyword and is browsed by the user;
the tag obtaining unit 3 is configured to obtain all classification tags corresponding to the clicked pages, and record the classification tags as first designated tags, where different classification tags correspond to different types of pages, and a same classification tag corresponds to multiple different similar pages;
a matching tag unit 4, configured to match the specified keyword with each of the first specified tags, so as to obtain a tag with a highest matching degree with the specified keyword, and mark the tag as a second specified tag;
a page finding unit 5, configured to find a first recommended page from the second specified tag;
and the recommendation page unit 6 is used for recommending the first recommendation page to the specified user.
The classification recommendation device based on the search history is applied to various websites with search engines, such as Taobao, Baidu, Google and the like. In this embodiment, the specified user refers to a user who currently inputs a keyword in a search engine, and the specified keyword may be a keyword currently input in the search engine, for example, a phrase or a sentence such as "jeans", "congratulations for getting rich", or "white rabbit".
As described in the key receiving unit 1 and the traversal recording unit 2, after receiving the specified keyword input by the specified user, traversing the search history records of all users in the database, finding out the user who has input the specified keyword, and according to the click result of the specified keyword, specifically, the page that is obtained by searching according to the specified keyword and is browsed by the user.
For example, if the specified user inputs "jeans", the database is traversed by all users to obtain a plurality of different users who have searched for "jeans", and the recommended pages obtained by searching for "jeans" include a "purchase page for purchasing jeans", "teaching page for wearing with jeans matching", "introduction page for introducing jeans brand", and the like.
As described in the tag obtaining unit 3, the classification tags corresponding to all clicked pages are obtained, and in order to facilitate distinguishing the first designated tag, in this embodiment, the classification tags are configured for all pages, a tag system is established, different classification tags correspond to different types of pages, for example, types including shopping, teaching, introduction, and the like, the same classification tag corresponds to a plurality of different similar pages, for example, a "purchase page for buying jeans" and a "purchase page for buying congratulatory fortune charms" are different similar pages in the same classification tag "shopping", and each classification tag carries corresponding tag information for matching with a keyword.
Further, the classification labels may be classified into classes, such as a first class classification, a second class classification, a third class classification, a first class classification of "shopping", a second class classification of "purchasing clothes", "purchasing electronic products", a third class classification of "purchasing trousers", "purchasing jacket", and the like.
As described in the matching tag unit 4, after the first designated tag is found, the designated keyword is matched with the first designated tag, so as to find the tag with the highest matching degree with the designated keyword, specifically, the designated keyword is matched with the tag information of the first designated tag, and the matching can be performed in a combination of multiple matching manners, or one matching manner is directly adopted for matching, and after the tag with the highest matching degree is obtained by matching, the tag is marked as the second designated tag.
If the page finding unit 5 and the page recommending unit 6 find the first recommended page from the second designated tag, and recommend the recommended page to the designated user, it should be known that the second designated tag includes a plurality of clicked pages, and the first recommended page can be found through the click rate or according to the user attribute setting rule, and then recommended to the designated user. Specifically, in one embodiment, the finding the page unit 5 includes:
and the click finding subunit is used for finding the click page with the highest click frequency from the second specified label and recording the click page as the first recommended page.
In this embodiment, the second designated tab is associated with a plurality of different pages, which include the above clicked pages, and the clicked page with the highest click frequency, that is, the above first recommended page, can be found from the clicked pages, and it is noted that the click frequency here refers to the click of the user based on the input of the designated keyword search, rather than the click based on the input of other keywords search, for example, the designated keyword is "jeans", and after matching the classification tag of the shopping, the total click of "purchase page of jeans" is 100 times, and in a certain time, the click of "purchase page of jeans" is 60 times based on the input of the designated keyword "jeans", the click of "purchase page of jeans" is 40 times based on the input of the "jeans" keyword, and the total click of "purchase page of jeans" is 150 times, and the click of "is 50 times based on the input of the designated keyword of" jeans ", the click rate is calculated to be the highest and is still the "purchase page for buying jeans" here, even though the total click rate of the "purchase page for buying jeans" is larger than the total click rate of the "purchase page for buying jeans" based on the "shorts" keyword for 100 clicks.
In another embodiment, to improve the matching rate with the specified user, the finding the first recommended page may be performed based on an attribute of the specified user, and specifically, the finding the page unit 5 includes:
an acquisition record subunit, configured to acquire a search history record of the specified user;
the extraction characteristic subunit is used for extracting the user characteristics of the specified user from the search history record;
the calculation model subunit is used for inputting the user characteristics into a preset prediction model for calculation to obtain a prediction result, and the prediction model is a neural network model obtained by training by taking the user characteristics and a corresponding page as a training set;
a determination result subunit, configured to determine the first recommended page from the second specified tag according to the prediction result.
In the embodiment, a search history of a specified user is obtained, wherein the search history includes search keywords and browsed pages, user characteristics of the specified user are extracted by counting the search keywords, the types of the browsed pages and corresponding search times, the user characteristics are used for representing search preferences of the specified user, the user characteristics are input into a preset prediction model to be calculated to obtain a prediction result, the prediction result is the probability that the specified user may click each page, and the page with the maximum probability is used as a first recommended page. The prediction module is a neural network model formed by deep learning and training through a training set, wherein samples in the training set comprise user characteristics and pages which are most likely to be clicked by a user, and a specific training process is the prior art and is not repeated here.
In another embodiment, other clicked pages in the second designated tag can be ranked from high to low according to the probability, and then the clicked page with the preset ranking rank is selected as a recommended page and recommended to the designated user.
According to the classification recommendation device based on the search history provided by the invention, when the input specified keyword is received, the specified keyword searched by all users and the classification label of the click result are traversed, the classification label with the highest matching degree is found, and the page with the highest frequency in the classification label is recommended to the user, so that the search efficiency of the user is improved, the content displayed to the user can be ensured to be the content which can be clicked by the user as much as possible, and the requirements of the user are met.
In an embodiment, the search history based classification recommendation apparatus includes:
and the repeated traversal unit is used for obtaining the label with the second highest matching degree with the specified keyword, recording the label as a third specified label, finding the clicked page with the highest clicking frequency from the third specified label, recording the clicked page as a second recommended page, and so on, repeatedly traversing all the first specified labels until finding the clicked page with the highest clicking frequency in each first specified label to obtain a plurality of recommended pages, and recommending the recommended pages to the specified user.
In this embodiment, in order to further improve the search time of the user, find the content required by the user as much as possible, recommend the page with the highest click frequency in each classification tag as a recommended page to the specified user, specifically, after finding the first recommended page, obtain the tag with the second highest matching degree with the specified keyword again, then find the clicked page with the highest click frequency from the obtained clicked pages, and so on, obtain the tags with the third highest matching degree with the specified keyword respectively, and obtain the recommended page of each first specified tag until all the first specified tags are traversed by the tag with the fourth highest matching degree with the specified keyword, but certainly, the recommended page of each first specified tag may be searched without searching in the order of the matching degrees, and only all the first specified tags need to be traversed.
In another embodiment, the search history based classification recommendation apparatus includes:
the high-label obtaining unit is used for eliminating the classification labels of the currently found recommended pages, matching the specified keywords with the rest of the first specified labels, and obtaining the classification label with the highest matching degree with the specified keyword in the rest of the first specified labels;
the finding recommendation unit is used for finding a click page with the highest click frequency from the classification label with the highest matching degree with the specified keyword, and marking the click page as a third recommendation page;
and the recommendation page unit is used for repeatedly executing the steps of obtaining the high-grade unit and finding the recommendation unit to obtain a plurality of third recommendation pages and recommending the third recommendation pages to the specified user.
In this embodiment, for more accurate matching, after a corresponding recommended page is found in each classification tag, the classification tag may be removed, and then re-matching is performed, specifically, after the recommended page is found through the first specified tag, the first specified tag is removed, then, the remaining first specified tags are matched with the specified keywords again, the classification tag with the highest matching degree is obtained again, so that a third recommended page is obtained, the above steps are repeated until all the first specified tags are matched, a plurality of third recommended pages are obtained, sorting may be performed according to the order of obtaining the recommended pages, and then, the recommended pages sorted within the preset ranking are selected and recommended to the user.
In one embodiment, the matching tag unit 4 includes:
the information obtaining subunit is used for obtaining page information of all clicked pages in the first specified label;
the abstract sub-unit is used for extracting abstract information from the page information and generating label information by the abstract information and the label name of the first specified label;
the information matching subunit is used for matching the specified keywords with the label information of each first specified label to obtain a corresponding matching degree;
and the label sorting subunit is used for sorting the first designated labels from high to low according to the matching degree to obtain second designated labels with the highest matching degree.
In this embodiment, in the step S4, the second specified label with the highest matching degree is obtained by matching the specified keyword with the first specified label, specifically, the page information of all clicked pages in the first specified label is obtained first, then extracting corresponding features from all page information, assembling all the features into abstract information, and generating label information by the abstract information and the label name of the first appointed label, of course, in order to further improve matching precision, information such as label attribute can be added, then matching the appointed keywords with the tag information to obtain the matching degree of each first appointed tag and the appointed keywords, then sequencing according to the matching degree from high to ground to obtain a second appointed tag with the highest matching degree, therefore, the matching precision of the keywords and the page is further improved by adding the information related to the page into the tag information.
Referring to fig. 3, the present application further provides a computer-readable storage medium 10, in which a computer program 20 is stored in the storage medium 10, and when the computer program runs on a computer, the computer is caused to execute the classification recommendation method based on the search history described in the above embodiment.
Referring to fig. 4, the present application further provides a computer device 40 containing instructions, the computer device includes a memory 30 and a processor 50, the memory 30 stores a computer program 20, and the processor 30 executes the computer program 20 to implement the classification recommendation method based on the search history as described in the above embodiment.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product.
The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the application to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can store or a data storage device, such as a server, a data center, etc., that is integrated with one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are also included in the scope of the present application.
Claims (10)
1. A classification recommendation method based on search history records is characterized by comprising the following steps:
s1: receiving a specified keyword input by a specified user;
s2: traversing the search history records of all users in a database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is searched according to the specified keyword and is browsed by the user;
s3: obtaining classification tags corresponding to all the clicked pages, recording the classification tags as first designated tags, wherein different classification tags correspond to different types of pages, and the same classification tag corresponds to a plurality of different similar pages;
s4: matching the appointed keywords with each first appointed label to obtain a label with the highest matching degree with the appointed keywords, and marking as a second appointed label;
s5: finding a first recommended page from the second specified tag;
s6: and recommending the first recommendation page to the specified user.
2. The search history based classification recommendation method according to claim 1, wherein the step S5 includes:
s50: and searching the clicked page with the highest click frequency from the second appointed label, and recording the clicked page as the first recommended page.
3. The search history based classification recommendation method according to claim 1, wherein after the step S4, the method comprises:
s40: and obtaining a label with the second highest matching degree with the specified keyword, recording the label as a third specified label, finding the clicked page with the highest clicking frequency from the third specified label, recording the clicked page as a second recommended page, repeating the traversing of all the first specified labels until the clicked page with the highest clicking frequency in each first specified label is found, obtaining a plurality of recommended pages, and recommending the recommended pages to the specified user.
4. The search history based classification recommendation method according to claim 1, wherein after the step S6, the method comprises:
s7: after the classification tags of the currently found recommended pages are removed, matching the specified keywords with the rest of the first specified tags to obtain the classification tags with the highest matching degree with the specified keywords in the rest of the first specified tags;
s8: finding a click page with the highest click frequency from the classification label with the highest matching degree with the specified keyword, and recording the click page as a third recommendation page;
s9: and repeating the steps S7-S8 to obtain a plurality of third recommendation pages, and recommending the third recommendation pages to the specified user.
5. The search history based classification recommendation method according to claim 1, wherein the step S5 includes:
s51: acquiring a search history of the specified user;
s52: extracting the user characteristics of the specified user from the search history record;
s53: inputting the user characteristics into a preset prediction model for calculation to obtain a prediction result, wherein the prediction model is a neural network model obtained by training by taking the user characteristics and a corresponding page as a training set;
s54: and determining the first recommended page from the second designated label according to the prediction result.
6. The search history based classification recommendation method according to claim 1, wherein the step S4 includes:
s41: acquiring page information of all clicked pages in the first appointed label;
s42: extracting abstract information from the page information, and generating label information by the abstract information and the label name of the first specified label;
s43: matching the appointed keywords with the label information of each first appointed label to obtain a corresponding matching degree;
s44: and sorting the first appointed labels from high to low according to the matching degree to obtain the second appointed label with the highest matching degree.
7. The classification recommendation device based on the search history record is characterized by comprising:
a receiving key unit for receiving a specified keyword input by a specified user;
the traversal recording unit is used for traversing the search history records of all users in the database to acquire a click page corresponding to the specified keyword, wherein the click page is a page which is searched according to the specified keyword and is browsed by the user;
the system comprises a tag obtaining unit, a tag obtaining unit and a tag selecting unit, wherein the tag obtaining unit is used for obtaining classification tags corresponding to all clicked pages and recording the classification tags as first designated tags, different classification tags correspond to different types of pages, and the same classification tag corresponds to a plurality of different similar pages;
a matching label unit, configured to match the specified keyword with each of the first specified labels, so as to obtain a label with a highest matching degree with the specified keyword, and mark the label as a second specified label;
the page finding unit is used for finding a first recommended page from the second specified label;
and the recommendation page unit is used for recommending the first recommendation page to the specified user.
8. The search history based category recommendation device of claim 7, wherein the find page unit comprises:
and the click finding subunit is used for finding the click page with the highest click frequency from the second specified label and recording the click page as the first recommended page.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor when executing the computer program performs the steps of the search history based classification recommendation method of any one of claims 1 to 6.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the search history based classification recommendation method of any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110398716.1A CN113127736A (en) | 2021-04-14 | 2021-04-14 | Classification recommendation method and device based on search history |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110398716.1A CN113127736A (en) | 2021-04-14 | 2021-04-14 | Classification recommendation method and device based on search history |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113127736A true CN113127736A (en) | 2021-07-16 |
Family
ID=76776218
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110398716.1A Pending CN113127736A (en) | 2021-04-14 | 2021-04-14 | Classification recommendation method and device based on search history |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113127736A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113706260A (en) * | 2021-09-01 | 2021-11-26 | 镇江纵陌阡横信息科技有限公司 | E-commerce platform commodity recommendation method and device based on search content |
CN117688250A (en) * | 2024-02-04 | 2024-03-12 | 国网湖北省电力有限公司信息通信公司 | Unified data dynamic service management system and method suitable for electric power full scene |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102436495A (en) * | 2011-11-14 | 2012-05-02 | 百度在线网络技术(北京)有限公司 | Method and device for providing dynamic searching page |
CN105701155A (en) * | 2015-12-30 | 2016-06-22 | 百度在线网络技术(北京)有限公司 | Information push method and the device |
CN106445947A (en) * | 2015-08-06 | 2017-02-22 | 阿里巴巴集团控股有限公司 | Data searching method and system |
CN111177523A (en) * | 2018-11-09 | 2020-05-19 | 阿里巴巴集团控股有限公司 | Webpage recommendation method and device, electronic equipment and computer readable medium |
CN111859091A (en) * | 2020-07-21 | 2020-10-30 | 山东省科院易达科技咨询有限公司 | Search result aggregation method and device based on artificial intelligence |
-
2021
- 2021-04-14 CN CN202110398716.1A patent/CN113127736A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102436495A (en) * | 2011-11-14 | 2012-05-02 | 百度在线网络技术(北京)有限公司 | Method and device for providing dynamic searching page |
CN106445947A (en) * | 2015-08-06 | 2017-02-22 | 阿里巴巴集团控股有限公司 | Data searching method and system |
CN105701155A (en) * | 2015-12-30 | 2016-06-22 | 百度在线网络技术(北京)有限公司 | Information push method and the device |
CN111177523A (en) * | 2018-11-09 | 2020-05-19 | 阿里巴巴集团控股有限公司 | Webpage recommendation method and device, electronic equipment and computer readable medium |
CN111859091A (en) * | 2020-07-21 | 2020-10-30 | 山东省科院易达科技咨询有限公司 | Search result aggregation method and device based on artificial intelligence |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113706260A (en) * | 2021-09-01 | 2021-11-26 | 镇江纵陌阡横信息科技有限公司 | E-commerce platform commodity recommendation method and device based on search content |
CN117688250A (en) * | 2024-02-04 | 2024-03-12 | 国网湖北省电力有限公司信息通信公司 | Unified data dynamic service management system and method suitable for electric power full scene |
CN117688250B (en) * | 2024-02-04 | 2024-04-16 | 国网湖北省电力有限公司信息通信公司 | Unified data dynamic service management system and method suitable for electric power full scene |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9489401B1 (en) | Methods and systems for object recognition | |
CN110427563B (en) | Professional field system cold start recommendation method based on knowledge graph | |
US9678989B2 (en) | System and method for use of images with recognition analysis | |
US9058394B2 (en) | Matching and recommending relevant videos and media to individual search engine results | |
CN108763321B (en) | Related entity recommendation method based on large-scale related entity network | |
CN111046221B (en) | Song recommendation method, device, terminal equipment and storage medium | |
CN111191122A (en) | Learning resource recommendation system based on user portrait | |
EP1995669A1 (en) | Ontology-content-based filtering method for personalized newspapers | |
US11042594B2 (en) | Artificial intelligence for product data extraction | |
CN111324771B (en) | Video tag determination method and device, electronic equipment and storage medium | |
CN114238573B (en) | Text countercheck sample-based information pushing method and device | |
CN106164889A (en) | System and method for internal storage data library searching | |
CN102314452B (en) | A kind of method and system of being undertaken navigating by input method platform | |
CN110321437B (en) | Corpus data processing method and device, electronic equipment and medium | |
WO2021112984A1 (en) | Feature and context based search result generation | |
CN112380457A (en) | Accurate personalized recommendation method based on purchase information | |
CN113127736A (en) | Classification recommendation method and device based on search history | |
CN106407316B (en) | Software question and answer recommendation method and device based on topic model | |
CN105653547A (en) | Method and device for extracting keywords of text | |
CN113065018A (en) | Audio and video index library creating and retrieving method and device and electronic equipment | |
CN110647504B (en) | Method and device for searching judicial documents | |
EP3731108A1 (en) | Search system, search method, and program | |
US11507593B2 (en) | System and method for generating queryeable structured document from an unstructured document using machine learning | |
JP2006023968A (en) | Unique expression extracting method and device and program to be used for the same | |
Mohtaj et al. | Maglet: A Persian journal recommender system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210716 |
|
WD01 | Invention patent application deemed withdrawn after publication |