CN113343105A - Keyword page management system, method, equipment and computer readable storage medium - Google Patents
Keyword page management system, method, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN113343105A CN113343105A CN202110728195.1A CN202110728195A CN113343105A CN 113343105 A CN113343105 A CN 113343105A CN 202110728195 A CN202110728195 A CN 202110728195A CN 113343105 A CN113343105 A CN 113343105A
- Authority
- CN
- China
- Prior art keywords
- commodity
- keyword
- feature
- feature classification
- search
- 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
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
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9538—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The application relates to the technical field of in-station search, in particular to a keyword page generated according to input in-station search keywords, which comprises the following steps: receiving search keywords in a station, matching commodity information and obtaining a search result set related to commodities; classifying the commodity attributes of the search results to obtain a classification result set; performing feature recognition on each classification result set to obtain a feature classification subset; and determining the weight of the feature classification subset according to the historical access quantity, and generating a keyword page according to the weight ratio of the feature classification subset. The method and the device for displaying the search results overcome the technical problems that the keyword page generated by directly inputting the keywords is low in displayed result relevance and random in displayed results, and can help the user generate a more accurate search result page.
Description
Technical Field
The present application relates to the field of in-site search technologies, and in particular, to a keyword page management system, method, device, and computer-readable storage medium.
Background
The keywords of the website are important contents which bring traffic to the website and benefit to enterprises. The keyword pages generated by the website topics are used, and the conversion rate is improved through SEO (Search _ Engine _ Optimization), which often drives the ranking of the website to be increased, thereby bringing traffic to the website and improving the sales volume of the website.
The keyword page can provide a customer with a more definite shopping direction, the customer has keywords belonging to commodities that the customer wants to buy when shopping, and the keyword page can enable the customer to find out commodities with eyes from a website more quickly, so that a more friendly shopping experience is provided for the customer.
However, since the keyword page generated by directly inputting the keywords has a low relevance of the displayed result and a random displayed result, how to help the user generate a more accurate search result page is a technical problem to be solved urgently.
Disclosure of Invention
In order to improve the relevance of a keyword page display result generated by a keyword and generate a more accurate search result page, the application provides a keyword page management system, a keyword page management method, a keyword page management device and a computer-readable storage medium.
In a first aspect, the present application provides a keyword page management method, which adopts the following technical scheme:
a keyword page management method comprises the following steps:
receiving search keywords in a station, matching commodity information and obtaining a search result set related to commodities;
classifying the commodity attributes of the search results to obtain a classification result set;
carrying out feature recognition on the commodity feature set of each classification result to obtain a feature classification subset;
acquiring historical access data, and determining the weight of the feature classification subset according to the historical access data;
and generating a keyword page according to the weight ratio of the feature classification subset.
The commodity information with higher relevance is divided by classifying the commodity attributes of the search results, the accuracy of the search results is further improved according to the characteristic identification, and the commodity information is ranked according to the weight proportion, so that the commodity information with more access numbers is ranked more forward, and most of people can find the desired commodity. The method and the device have the effects of improving the relevancy of the keyword page display result generated by the keyword and generating a more accurate search result page.
Optionally, a method for managing a keyword page, where the keyword page includes a main product list and a page data information list, and the generating a keyword page according to the weight proportion of the feature classification subset includes the following steps:
when the weight proportion of the feature classification subset is greater than a first threshold value, the feature classification subset is in the expanded main commodity list information and is displayed in the main commodity list according to the feature classification subset;
when the weight proportion of the feature classification subset is between the first threshold and the second threshold, the feature classification subset is in the folded main commodity list information, corresponding feature category data are generated according to the feature classification subset, and the feature classification subset is displayed in the page data information list;
wherein the second threshold is less than the first threshold.
By adopting the technical scheme, the weight of the feature classification subset is compared.
Optionally, the method for managing a keyword page further includes, before searching for a keyword in the receiving station, establishing an original database, where the original database includes a plurality of commodity information, and each commodity information includes a commodity keyword, a commodity attribute, and a commodity feature.
By adopting the technical scheme, when the weight proportion of the feature classification subset is greater than a first threshold value, the historical access data volume is high, and the content with the high historical access data volume is preferentially displayed in the main commodity list; when the historical access data volume is low, the feature classification subset is folded and is not displayed in the main commodity list, but feature classification data is generated and displayed in the page data information list and can be displayed in the main commodity list only according to access request instructions and other modes; when the amount of historical access data is low, indicating that the part is not a satisfactory search result, the feature classification subset of the part will not be displayed. The search results meeting the requirements are preferentially displayed by dividing the weight intervals of the feature classification subsets, so that the relevancy of the results displayed on the keyword page is improved.
Optionally, a keyword page management method, where the specific method for collecting historical access data includes:
acquiring the time length from commodity information access to ordering of a user, comparing the time length with a time preset value, judging that the time length is abnormal if the time length is less than the time preset value, and removing the related historical access data according to the abnormal time length.
By adopting the technical scheme, whether the seller has a single brushing line or not is judged by judging the time length from the commodity information access to the order placing of the user, and the general time length of the seller having the single brushing behavior is shorter, so that if the time length is less than the time preset value, the historical access data can be considered invalid, the historical access data can be removed, and the judgment of the weight of the feature classification subset is more accurate.
Optionally, a keyword page management method, where the specific method for collecting historical access data includes:
obtaining comment pictures of a user on a single commodity, identifying all the comment pictures of the single commodity, judging the similarity between the comment pictures, judging that the comment pictures are abnormal if the similarity of at least two comment pictures is larger than a duplication checking index threshold value, and removing related historical access data according to the abnormal comment pictures.
By adopting the technical scheme, whether the seller has the single file swiping behavior or not is judged by judging the similarity between the comment pictures, the seller who usually has the single file swiping behavior can steal or repeatedly use the same comment picture, and the comment picture can be slightly modified and then steal or repeatedly use the picture, so that if the similarity of the comment picture is greater than the duplication checking index threshold value, the history access data can be considered invalid, the history access data can be removed, and the judgment of the weight of the feature classification subset is more accurate.
Optionally, a keyword page management method, where the specific method for determining similarity between comment pictures includes:
inputting all the comment pictures of a single commodity into an identification detection model, and matching the similarity between the comment pictures;
and if the same elements among the comment pictures exceed the duplicate checking index threshold value, judging that the comment pictures are abnormal.
By adopting the technical scheme, whether the comment pictures are abnormal or not is judged according to the judgment whether the same elements among the comment pictures exceed the duplicate checking index threshold value or not, and whether the behavior of the seller for swiping the bill exists in the historical data or not can be accurately judged.
Optionally, a keyword page management method, where the specific method for feature recognition includes:
and acquiring the commodity picture to be identified in each classification result set, and putting the commodity picture to be identified into a convolutional neural network model obtained by pre-training to obtain a color matching class set and/or a component class set of the commodity picture.
By adopting the technical scheme, the commodity pictures are identified in the classification result set, and the commodities in the classification result set are classified according to the color matching or component category of the commodity pictures, so that the dimensionality of a search result page can be increased, and a user can be helped to generate a more accurate search result page.
In a second aspect, the present application provides a keyword page management system, which adopts the following technical solutions:
a keyword page management system, comprising:
the search module is used for receiving the in-station search keywords, matching the commodity information and obtaining a search result set related to the commodity;
the classification module is used for classifying the commodity attributes of the search results to obtain a classification result set;
the identification module is used for carrying out feature identification on each classification result set to obtain a feature classification subset;
and the page generation module is used for determining the weight of the feature classification subset according to the historical access quantity and generating a keyword page according to the weight ratio of the feature classification subset.
By adopting the technical scheme, the commodity attribute classification is carried out on the search results, the commodity information with higher relevance is divided, the accuracy of the search results is further improved according to the feature identification, and the commodity information is ranked according to the weight proportion, so that the commodity information rows with more access numbers are more forward, and most of people can find the desired commodities. The method and the device have the effects of improving the relevancy of the keyword page display result generated by the keyword and generating a more accurate search result page.
In a third aspect, the present application provides a keyword page management device, which adopts the following technical solution:
a keyword page management apparatus comprising a memory and a processor, the processor being capable of being loaded and executing a keyword page management method as claimed in any one of the preceding claims.
In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solutions:
a computer readable storage medium storing a keyword page management method as claimed in any one of the above claims, which can be loaded and executed by a processor.
Detailed Description
The embodiment of the application discloses a keyword page management method.
A keyword page management method comprises the following steps:
s0, establishing an original database, and inputting commodity information of the commodities into the original database, so that the original database has a plurality of commodity information, and each commodity information comprises a commodity keyword, a commodity attribute and a commodity characteristic.
In one embodiment, if the commodity is a notebook computer, the commodity keyword may be "notebook", "super-notebook", "entertainment book", "game book", and the keyword may also be associated with a name, a series name, a memory, a size, a CPU series, etc. of a corresponding brand, for example, the keyword may be "new notebook MateBook 24 inch 2K high definition eight-core standard-pressure commerce game i7 fingerprint metal 2024 portable computer"; the commodity attribute is 'notebook computer'; the commodity features may include brand "medium" and "24" in size, "512G in hard disk capacity," 16G "in memory," independent display card "in display card type," "fingerprint identification," "face unlock," "wireless charging," and so on.
And S1, receiving the in-station search keywords, matching the commodity information, and obtaining a search result set related to the commodity.
In a specific embodiment, after receiving the in-station search keyword "notebook", the in-station search keyword "notebook" is paired with a large amount of commodity information in the database, and all search results containing the in-station search keyword "notebook" are generated.
And S2, classifying the commodity attributes of the search results to obtain a classification result set.
According to the commodity attributes of the search results, the search results can be classified into a notebook computer and a notebook stationery, irrelevant search results can be removed by classifying the commodity attributes, the commodity search range can be narrowed, and the commodity search relevance can be improved.
And S3, performing feature recognition on the commodity features of each classification result set to obtain a feature classification subset. In a specific embodiment, the feature classification can be classified into "16G", "8G", "4G", and the like from the memory dimension, the display card type can be classified into "independent display card", "integrated display card", and a notebook can belong to the classification of "16G", "independent display card".
The specific method for identifying the characteristics may further include obtaining a commodity picture to be identified in each classification result set, and putting the commodity picture to be identified into a convolutional neural network model obtained through pre-training to obtain at least one of a color matching category set and a component category set of the commodity picture. Since the convolutional neural network model is a prior art, it is not described in detail herein.
And S4, collecting historical access data, and determining the weight of the feature classification subset according to the historical access data. The specific method for collecting the historical access data comprises the following steps:
and acquiring the time-consuming length from the commodity information access of the user to the ordering, comparing the time-consuming length with a time preset value, judging that the time-consuming length is abnormal if the time-consuming length is less than the time preset value, and removing related historical access data according to the abnormal time-consuming length. The use length of the seller who usually has the list-swiping behavior is short, so if the use length is less than the time preset value, the historical access data can be considered to be invalid, and the historical access data can be eliminated. Whether the seller has a brushing line function or not is judged by judging the time length from the commodity information access to the order placing of the user, so that the judgment of the weight of the feature classification subset is more accurate.
The specific method for acquiring the historical access data comprises the steps of obtaining comment pictures of a user on a single commodity, identifying all the comment pictures under the single commodity, judging the similarity between the comment pictures, judging that the comment pictures are abnormal if the similarity of at least two comment pictures is larger than a duplication index threshold value, and removing the related historical access data according to the abnormal comment pictures.
The specific method for judging the similarity between the comment pictures comprises the steps of inputting all the comment pictures of a single commodity into an identification detection model, and matching the similarity between the comment pictures;
and if the same elements among the comment pictures exceed the duplicate checking index threshold value, judging that the comment pictures are abnormal.
And S5, generating a keyword page according to the weight ratio of the feature classification subset.
The keyword page comprises a main commodity list and a page data information list, and the keyword page is generated according to the weight proportion of the feature classification subset comprises the following steps: and when the weight ratio of the feature classification subset is greater than a first threshold value, the feature classification subset is in the expanded main commodity list information and is displayed in the main commodity list according to the feature classification subset. And when the weight ratio of the feature classification subset is between a first threshold and a second threshold, the feature classification subset is in the folded main commodity list information, corresponding feature category data are generated according to the feature classification subset, and the feature category data are displayed in a page data information list. Wherein the second threshold is less than the first threshold.
The implementation principle of the keyword page management method in the embodiment of the application is as follows: the commodity information with higher relevance is divided by classifying the commodity attributes of the search results, the accuracy of the search results is further improved according to the characteristic identification, and the commodity information is ranked according to the weight proportion, so that the commodity information with more access numbers is ranked more forward, and most of people can find the desired commodity. The method and the device have the effects of improving the relevancy of the keyword page display result generated by the keyword and generating a more accurate search result page.
A keyword page management system, comprising:
the search module is used for receiving the in-station search keywords, matching the commodity information and obtaining a search result set related to the commodity;
the classification module is used for classifying the commodity attributes of the search results to obtain a classification result set;
the identification module is used for carrying out feature identification on each classification result set to obtain a feature classification subset;
and the page generation module is used for determining the weight of the feature classification subset according to the historical access quantity and generating a keyword page according to the weight ratio of the feature classification subset.
The implementation principle of the keyword page management method in the embodiment of the application is as follows: the commodity information with higher relevance is divided by classifying the commodity attributes of the search results, the accuracy of the search results is further improved according to the characteristic identification, and the commodity information is ranked according to the weight proportion, so that the commodity information with more access numbers is ranked more forward, and most of people can find the desired commodity. The method and the device have the effects of improving the relevancy of the keyword page display result generated by the keyword and generating a more accurate search result page.
The above embodiments are preferred embodiments of the present application, and the protection scope of the present application is not limited by the above embodiments, so: all equivalent changes made according to the structure, shape and principle of the present application shall be covered by the protection scope of the present application.
Claims (10)
1. A keyword page management method is characterized by comprising the following steps:
receiving search keywords in a station, matching commodity information and obtaining a search result set related to commodities;
classifying the commodity attributes of the search results to obtain a classification result set;
carrying out feature recognition on the commodity feature set of each classification result to obtain a feature classification subset;
acquiring historical access data, and determining the weight of the feature classification subset according to the historical access data;
and generating a keyword page according to the weight ratio of the feature classification subset.
2. The method for managing the keyword page according to claim 1, wherein the keyword page comprises a main commodity list and a page data information list, and the step of generating the keyword page according to the weight ratio of the feature classification subset comprises the following steps:
when the weight proportion of the feature classification subset is greater than a first threshold value, the feature classification subset is in the expanded main commodity list information and is displayed in the main commodity list according to the feature classification subset;
when the weight proportion of the feature classification subset is between the first threshold and the second threshold, the feature classification subset is in the folded main commodity list information, corresponding feature category data are generated according to the feature classification subset, and the feature classification subset is displayed in the page data information list;
wherein the second threshold is less than the first threshold.
3. The method as claimed in claim 1, further comprising building an original database before searching for the keyword in the receiving station, wherein the original database contains a plurality of commodity information, and each commodity information comprises a commodity keyword, a commodity attribute and a commodity feature.
4. The method for managing the keyword pages as claimed in claim 3, wherein the specific method for collecting the historical access data comprises:
acquiring the time length from commodity information access to ordering of a user, comparing the time length with a time preset value, judging that the time length is abnormal if the time length is less than the time preset value, and removing the related historical access data according to the abnormal time length.
5. The method for managing the keyword pages as claimed in claim 3, wherein the specific method for collecting the historical access data comprises:
obtaining comment pictures of a user on a single commodity, identifying all the comment pictures of the single commodity, judging the similarity between the comment pictures, judging that the comment pictures are abnormal if the similarity of at least two comment pictures is larger than a duplication checking index threshold value, and removing related historical access data according to the abnormal comment pictures.
6. The method for managing the keyword pages as claimed in claim 3, wherein the specific method for judging the similarity between the comment pictures comprises the following steps:
inputting all the comment pictures of a single commodity into an identification detection model, and matching the similarity between the comment pictures;
and if the same elements among the comment pictures exceed the duplicate checking index threshold value, judging that the comment pictures are abnormal.
7. The method for managing the keyword page according to claim 1, wherein the specific method for identifying the features comprises the following steps:
and acquiring the commodity picture to be identified in each classification result set, and putting the commodity picture to be identified into a convolutional neural network model obtained by pre-training to obtain a color matching class set and/or a component class set of the commodity picture.
8. A keyword page management system, comprising:
the search module is used for receiving the in-station search keywords, matching the commodity information and obtaining a search result set related to the commodity;
the classification module is used for classifying the commodity attributes of the search results to obtain a classification result set;
the identification module is used for carrying out feature identification on each classification result set to obtain a feature classification subset;
and the page generation module is used for determining the weight of the feature classification subset according to the historical access quantity and generating a keyword page according to the weight ratio of the feature classification subset.
9. A keyword page management apparatus comprising a memory and a processor, wherein the processor is capable of being loaded and executing the keyword page management method according to any one of claims 1 to 6.
10. A computer-readable storage medium storing a keyword page management method according to any one of claims 1 to 6, which can be loaded and executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110728195.1A CN113343105A (en) | 2021-06-29 | 2021-06-29 | Keyword page management system, method, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110728195.1A CN113343105A (en) | 2021-06-29 | 2021-06-29 | Keyword page management system, method, equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113343105A true CN113343105A (en) | 2021-09-03 |
Family
ID=77481580
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110728195.1A Pending CN113343105A (en) | 2021-06-29 | 2021-06-29 | Keyword page management system, method, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113343105A (en) |
-
2021
- 2021-06-29 CN CN202110728195.1A patent/CN113343105A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11222055B2 (en) | System, computer-implemented method and computer program product for information retrieval | |
| US10354308B2 (en) | Distinguishing accessories from products for ranking search results | |
| JP5421737B2 (en) | Computer implementation method | |
| US9767144B2 (en) | Search system with query refinement | |
| US20060026152A1 (en) | Query-based snippet clustering for search result grouping | |
| US8606789B2 (en) | Method for layout based document zone querying | |
| US9720979B2 (en) | Method and system of identifying relevant content snippets that include additional information | |
| US20090319500A1 (en) | Scalable lookup-driven entity extraction from indexed document collections | |
| US20080040342A1 (en) | Data processing apparatus and methods | |
| CN103136228A (en) | Image search method and image search device | |
| JP6346218B2 (en) | Search method, apparatus and server for online trading platform | |
| CN110968800B (en) | Information recommendation method and device, electronic equipment and readable storage medium | |
| US8825641B2 (en) | Measuring duplication in search results | |
| US20100042610A1 (en) | Rank documents based on popularity of key metadata | |
| US10366108B2 (en) | Distributional alignment of sets | |
| Zhang et al. | A coarse-to-fine framework to efficiently thwart plagiarism | |
| US20120158716A1 (en) | Image object retrieval based on aggregation of visual annotations | |
| CN113449187A (en) | Product recommendation method, device and equipment based on double portraits and storage medium | |
| US20100138414A1 (en) | Methods and systems for associative search | |
| US20140105509A1 (en) | Systems and methods for comparing images | |
| US20140236939A1 (en) | Systems and methods for topical grouping of search results and organizing of search results | |
| KR20120038418A (en) | Searching methods and devices | |
| CN112487161A (en) | Enterprise demand oriented expert recommendation method, device, medium and equipment | |
| CN111160699A (en) | Expert recommendation method and system | |
| US11720626B1 (en) | Image keywords |
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 |