CN114493786A - Information recommendation method and device - Google Patents
Information recommendation method and device Download PDFInfo
- Publication number
- CN114493786A CN114493786A CN202210092529.5A CN202210092529A CN114493786A CN 114493786 A CN114493786 A CN 114493786A CN 202210092529 A CN202210092529 A CN 202210092529A CN 114493786 A CN114493786 A CN 114493786A
- Authority
- CN
- China
- Prior art keywords
- information
- item
- user
- recommended
- content
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses an information recommendation method and device, and relates to the technical field of computers. One embodiment of the method comprises: responding to the trigger of a user aiming at the target area in the display page, and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information; taking the first user characteristic and the plurality of item information as the input of a sequencing model, and determining the sequencing result of the plurality of item information according to the output of the sequencing model; the sequencing model is obtained by training according to historical behavior information corresponding to a plurality of second user characteristics and a plurality of second user characteristics respectively; determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregating recommended contents of a plurality of item information; and recommending the target information through a display page. According to the embodiment, the effectiveness and the efficiency of information recommendation are improved, and further the user experience is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to an information recommendation method and device.
Background
The e-commerce platform generally carries out marketing through coupons, kills of seconds, pre-sale and the like.
When a user browses a marketing message, the user often sees an item list corresponding to the marketing message. For example, taking a user browsing marketing information corresponding to a coupon as an example, the user sees a list of all item information associated with the coupon, the item information is randomly arranged in the list, and the user may need to perform operations such as page turning, searching again, and the like for many times to obtain content recommendation of interest of the user, so that the effectiveness of information recommendation is reduced, and the user experience is reduced.
Disclosure of Invention
In view of this, embodiments of the present invention provide an information recommendation method and apparatus, which can recommend item syndication content corresponding to a target area to a user through a display page according to a sorting result of a plurality of item information corresponding to the target area by a sorting model. Therefore, the operation times of the user are reduced, the interested content can be accurately recommended to the user, the effectiveness and efficiency of information recommendation are improved, and the user experience is further improved.
Furthermore, the information of a plurality of articles can be used as the input of the content generation model, the recommended content can be generated and aggregated for the information of the plurality of articles, the data volume of the recommended information is reduced, and the user can quickly acquire the recommended key points from the recommended information, so that the accuracy of the recommended information is improved. In addition, the sorting model is optimized according to the feedback information recommended by the user aiming at the information, the recommendation accuracy of the sorting model is improved, and therefore the click conversion rate of the articles is improved.
In order to achieve the above object, according to a first aspect of an embodiment of the present invention, there is provided an information recommendation method including:
responding to the trigger of a user for a target area in a display page, and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information;
taking the first user characteristic and the plurality of item information as input of a sorting model, and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively;
determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information;
and recommending the target information through the display page.
Optionally, before determining the target information to be recommended according to the sorting result and the item aggregation content, the method further includes:
the plurality of item information is used as the input of a content generation model to obtain the recommended content corresponding to the plurality of item information respectively; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information;
and aggregating the recommended contents to obtain the item aggregated contents.
Optionally, before receiving the trigger of the user for the target area in the presentation page, the method further includes:
obtaining recommendation target parameters, wherein the recommendation target parameters comprise any one or more of the following items: return on investment, inventory consumption parameters, total value and the number of people to be recommended;
and determining a plurality of item information corresponding to the target area according to the recommended target parameter, the type of the target area and the item information which can be displayed on the display page.
Optionally, the item information comprises an item category;
determining target article information from the article information corresponding to the article category; the target item information is item information of which the historical order quantity, the order completion quantity and/or the order value are/is larger than a preset threshold value;
and taking the target item information as an input of the content generation model.
Optionally, the method further comprises:
feedback information of the user aiming at the target information is obtained through the display page;
and optimizing the sequencing model according to the feedback information and preset time length.
Optionally, the ranking model is trained based on an LR algorithm, a GBDT algorithm, an Xgboost algorithm, a LightGBM algorithm, an xDeepFM algorithm, a DeepFM algorithm, and an AutoInt algorithm.
According to a second aspect of the embodiments of the present invention, there is provided an information recommendation apparatus including: the system comprises an acquisition module, a sorting module and a recommendation module; wherein the content of the first and second substances,
the acquisition module is used for responding to the trigger of a user aiming at a target area in a display page and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information;
the sorting module is used for taking the first user characteristic and the plurality of item information as input of a sorting model and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively;
the recommending module is used for determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information; and recommending the target information through the display page.
Optionally, the apparatus further comprises: a polymerization module; wherein the content of the first and second substances,
the aggregation module is configured to use the plurality of item information as an input of a content generation model to obtain the recommended content corresponding to each of the plurality of item information; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information; and aggregating the recommended contents to obtain the item aggregated contents.
According to a third aspect of embodiments of the present invention, there is provided an electronic apparatus, including:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method as in any one of the information recommendation methods provided in the first aspect above.
According to a fourth aspect of embodiments of the present invention, there is provided a computer-readable medium, on which a computer program is stored, which when executed by a processor, implements the method according to any one of the information recommendation methods provided in the above first aspect.
One embodiment of the above invention has the following advantages or benefits: according to the sorting result of the plurality of item information corresponding to the target area by the sorting model, the item aggregation content corresponding to the target area can be recommended to the user through the display page. Therefore, the operation times of the user are reduced, the interested content can be accurately recommended to the user, the effectiveness and efficiency of information recommendation are improved, and the user experience is further improved.
Furthermore, the information of a plurality of articles can be used as the input of the content generation model, the recommended content can be generated for the information of the plurality of articles quickly and aggregated, the data volume of the recommended information is reduced, so that the user can quickly acquire recommended key points from the recommended information, and the accuracy of the recommended information is improved. In addition, the sorting model is optimized according to the feedback information recommended by the user aiming at the information, so that the recommendation accuracy of the sorting model is improved, and the click conversion rate of the articles is improved.
Further effects of the above-mentioned non-conventional alternatives will be described below in connection with the embodiments.
Drawings
The drawings are included to provide a better understanding of the invention and are not to be construed as unduly limiting the invention. Wherein:
fig. 1 is a schematic flow chart of an information recommendation method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an information recommendation device for an e-commerce platform according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a ranking model according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating an information recommendation method for an e-commerce platform according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an information recommendation apparatus according to an embodiment of the present invention;
FIG. 6 is an exemplary system architecture diagram in which embodiments of the present invention may be applied;
fig. 7 is a schematic block diagram of a computer system suitable for use in implementing a terminal device or server of an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention are described below with reference to the accompanying drawings, in which various details of embodiments of the invention are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
It should be noted that embodiments of the present invention and technical features in the embodiments may be combined with each other without conflict.
As shown in fig. 1, an embodiment of the present invention provides an information recommendation method, which may include the following steps S101 to S104:
step S101: responding to the trigger of a user for a target area in a display page, and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information.
Step S102: taking the first user characteristic and the plurality of item information as input of a sorting model, and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively.
Step S103, determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregating the recommended contents of the plurality of item information.
And step S104, recommending the target information through the display page.
It will be appreciated that the presentation page may be a business system page viewable by the user. The target area can be a portal which can be triggered by a user, for example, in an e-commerce platform, the target area of the display page can be a content recommendation portal provided for the user by finding a page, planting a sketch page, shopping and the like, and can also be a portal of marketing information which can be viewed by the user, such as a coupon link, a killing page and a pre-sale page. It will be appreciated that a user may browse a plurality of item information by entering the corresponding page from these entries or pages. The browsable item information is a plurality of item information corresponding to the target areas corresponding to the inlets.
The information recommendation method provided by the embodiment of the invention can be applied to a display page of an e-commerce platform so as to recommend commodity information. When applied to an e-commerce platform, the information recommendation method provided by the embodiment of the invention can be implemented by the information recommendation device shown in fig. 2. As shown in fig. 2, the apparatus may include an application presentation module, a recommendation module, a data storage module, a marketing content generation module, and a marketing algorithm module.
In the step S101, the user may trigger the display page through a content entry or a marketing entry of the application display module, and accordingly, the content entry or the marketing entry is the corresponding target area, and the browsable item information corresponding to the content entry or the marketing entry is the plurality of item information corresponding to the target area. In step S102, a ranking model may be set in the recommending module to implement thousands of people and thousands of faces according to the ranking result of the ranking model. In addition, the recommended content and the item aggregated content in step S103 may be generated based on the marketing content generation module, and in generating the recommended content, the marketing content generation module may generate the recommended content corresponding to the marketing good based on the marketing good information and aggregate the recommended content and the marketing good into the good aggregated content. Specific embodiments of generating the recommended content and the item syndication content will be described in further detail in the examples hereinafter.
Before receiving the trigger of the user for the target area in the display page, the item information corresponding to the target area may be determined, and the following manner provided by an embodiment of the present invention may be adopted: acquiring recommendation target parameters, wherein the recommendation target parameters comprise any one or more of the following items: return on investment, inventory consumption parameters, total value and the number of people to be recommended; and determining a plurality of item information corresponding to the target area according to the recommended target parameter, the type of the target area and the item information which can be displayed on the display page.
The item information corresponding to the target area is a subset obtained from all the displayable item information associated with the display page according to the recommended target parameter and the type of the target area.
It will be appreciated that the recommendation target parameter may be determined by a combination of one or more parameters. The type of the target area can be tab page entry, link entry, picture entry, and the like.
When a marketing is created on the e-commerce platform, a marketing target is usually set for the marketing, which may be the return on investment of items participating in the marketing, the number of items expected to be consumed in inventory, the total amount of value to be realized, or the number of users facing the marketing, and the number of recommended users. Wherein, the total value can be determined according to GMV (Gross Merchandis Volume). When a marketing plan is created, the recommendation target parameters can be input into a marketing algorithm module to cause the marketing algorithm module to output corresponding item information. The marketing algorithm module provided by the embodiment of the invention can be as shown in fig. 2.
After determining the plurality of item information corresponding to the target area, the item aggregation content corresponding to the plurality of item information may be determined, and the following manner provided by an embodiment of the present invention may be adopted: the plurality of item information is used as the input of a content generation model to obtain the recommended content corresponding to the plurality of item information respectively; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information; and aggregating the recommended contents to obtain the item aggregated contents.
When applied to an e-commerce platform, the content generation model provided by the embodiment of the invention may be a marketing content generation module shown in fig. 2. The marketing content generating module may generate one or more recommended contents corresponding to the plurality of item information, that is, a recommended short title, a recommended short text, recommended case information, comment information, and recommended short video information, from the content data of the business system according to the plurality of input item information by using a natural language processing technology and an image processing technology. The comment information can include text comment information and picture comment information, and the picture comment information can be a sun note comment and the like.
It is to be understood that when the item information is an item category, the target item information that may be input as a content generation model may be determined in the following manner provided by one embodiment of the present invention: determining target article information from the article information corresponding to the article category; the target item information is item information of which the historical order quantity, the order completion quantity and/or the order value are/is larger than a preset threshold value.
For example, the total number of historical orders for the item category is M, the total number of completed orders is N, and the total number of order values is S. A preset threshold value M 'may be set to 80% × M, and in the item category, the item information whose historical order number is greater than M' is used as the target item information; a preset threshold value N 'may be set to 80% × N, and in the item category, the item information whose historical order completion amount is greater than N' may be used as the target item information; a preset threshold value S 'may be set to 80% S, and in the item category, item information having a historical order value greater than S' may be used as the target item information. The three target item information which is larger than the preset threshold value can be intersected to obtain final target item information, namely, the item information with the historical order quantity larger than M ', the historical order completion quantity larger than N ' and the historical order value larger than S ' is used as the target item information.
After the recommended content corresponding to each of the plurality of item information is obtained, the plurality of item information and the recommended content corresponding to the target area are aggregated to obtain an item aggregated content. Taking e-commerce platform as an example, the aggregation process can be expressed as:
wherein
Item information representing the m dimensions contained by the marketing k.
When the method and the device are applied to an e-commerce platform, the item aggregation content can be stored into a Redis database or a MYSQL database through the data storage module in FIG. 2, and can also be stored into an HBASE database or an Elasticissearch database.
After the plurality of item information in the target area is determined and the item aggregates corresponding to the plurality of item information are generated and stored, when a trigger of the user for the target area is received, one possible implementation manner of step S102 is to sort the plurality of item information for the user. The first user characteristic corresponding to the user can be obtained first, and meanwhile, historical behavior information corresponding to the first user characteristic can also be obtained. The first user characteristic may be personal information, order information, and the like of the user, and the corresponding historical behavior information may be content characteristics browsed by the user, item characteristics, cross characteristics, request scene characteristics, session characteristics, and the like of a plurality of items corresponding to the target area.
The article characteristics are detailed information of the article, such as article identification, article type, and the like. The cross feature is a feature with richer dimensions obtained by crossing different features. And the request scenario feature may be a geographic location feature where the user of the trigger target area is located. The session feature is a session feature that is saved when a user jumps between multiple sessions.
And inputting the plurality of item information corresponding to the target area, the first user characteristic and the user behavior information corresponding to the first user characteristic into a sequencing model for sequencing to obtain the sequence of recommending the plurality of item information to the user.
In one possible implementation manner of step S102, the ranking model is trained based on LR algorithm, GBDT algorithm, Xgboost algorithm, LightGBM algorithm, xDeepFM algorithm, deep fm algorithm, and AutoInt algorithm. The ranking model is the model used in the recommendation module of figure 2.
In one embodiment of the present invention, a schematic diagram of the structure of the order model may be as shown in FIG. 3. The feature input may be a user feature and user behavior information, that is, a user feature, a content feature, an article feature, a cross feature, a request scene feature, and a session feature. The training process of the ranking model can be obtained by training historical behavior information corresponding to a plurality of user characteristics and a plurality of user characteristics respectively.
A plurality of sub-models are used in the ranking model in fig. 3 to fit the ranking results to the content of interest to the user. Wherein, the characteristic input is preprocessed by using sub-models such as GBDT, LR, FM, Wide & Deep and the like. The cross-network, i.e., deep learning model, of the middle layer may include: xDeepFM, DeepFM, DCN, NFM, AFM, AutoInt, etc.; the characteristic characterization can use DSIN, DIEN, DIN and the like; the MMOE and ESSM can then be used as a model for multitask learning. The ranking model can combine LR algorithm, GBDT algorithm, Xgboost algorithm and LightGBM to construct a tree model for model training.
The deep learning models xDeepFM, DeepFM and AutoInt can well combine depth and breadth to mine high-order and low-order cross information for user behavior information, and the model learning efficiency is improved.
Understandably, machine learning models degrade very quickly when data changes. In one embodiment of the invention, a mode of combining the LR algorithm and the GBDT algorithm is used, so that the LR model can supplement the deficiency of online learning of the GBDT model, and the performance of the model is not reduced too much when online data changes. For example, in an e-commerce platform, input feature data is generally large, input data changes rapidly, and the online learning capability of the ranking model is strong, so that the result output by the ranking model can be ensured to be more accurate.
It can be understood that, usually, the business system is continuously acquiring a large amount of user characteristics and user behavior information, the ranking model usually needs to be updated and calibrated in real time, that is, the model needs to be retrained continuously by using real-time data, and the training speed of the ranking model and the data processing speed are critical at this time. Therefore, in an embodiment of the present invention, the LightGBM is adopted in the ranking model, which can support high-efficiency parallel training, so as to obtain the effects of higher training speed, lower memory consumption and fast processing of mass data, and improve the efficiency of information recommendation.
In addition, the Xgboost can reduce the possibility of overfitting of the sequencing model, so that the loss of the model is more accurate, when the training data is a sparse value, the default direction of a branch can be specified for a missing value or a specified value, and the training efficiency of the model can be greatly improved.
In an embodiment of the present invention, the target (target) of the output layer of the ranking model in fig. 3 is generally formed by fusing CTR (Click-Through-Rate), CVR (Conversion Rate), and GMV (Gross trade Volume), and the fusing manner of these targets may be generally performed by fusing exponentially or fusing by linear weight.
And a Re-Rank layer (Re-Rank layer) can be arranged on the output layer of the ranking model, and the results of the output layer can be Re-ranked according to the regional characteristics, diversity characteristics, user interest exploration proportions and the like of a plurality of articles, so that the final result of the ranking model is more in line with the content of interest of the user. The sequencing result of the sequencing model indicates the recommended sequence of the plurality of articles corresponding to the target area on the display page.
And displaying the item aggregation contents corresponding to the item information in the display page according to the sequencing result so as to recommend information to the user.
For example, in the e-commerce platform, the item A, B, C with the recommendation degree from high to low shown by the ranking result is ranked first by the item a, the aggregated content (content such as recommended documents, recommended short videos, comment information and the like) corresponding to the item a is displayed at the top of the display page, the aggregated content related to the item B is ranked second, and so on.
In the using process of the service system, the accuracy of the sequencing model can be determined according to the feedback information of the user on the target information, and the sequencing model is updated and calibrated in real time. The following may be employed as provided by one embodiment of the invention: feedback information of the user aiming at the target information is obtained through the display page; and optimizing the sequencing model according to the feedback information and preset time length.
It can be understood that the feedback information includes positive feedback information, for example, the user has a higher click rate on the top-ranked articles in the target information, a long browsing time, a higher number of sharing times, a higher order placing amount, and the like, in other words, the articles with higher recommendation degree have higher attention and conversion rate; the feedback information also includes negative feedback information, for example, the user does not click on the items ranked in the front, or directly closes the display page, or the like, in other words, the items with higher recommendation degrees do not accord with the content in which the user is interested.
And performing incremental training and calibration on the model according to the positive and negative feedback information. The preset duration can be in a day unit, namely, the sequencing model is subjected to incremental training by using historical data of the previous day every day to calibrate and update; the preset duration may also be in hours, i.e., the ranking model is incrementally trained to calibrate updates every hour using historical data from the previous hour. The accuracy of the sequencing model can be improved by continuously calibrating and updating the model, the possibility that the sequencing result accords with the content of interest of the user is improved, and the information recommendation accuracy is further improved.
The following describes the information recommendation method provided by the embodiment of the present invention in detail by taking an information recommendation method applied to an e-commerce platform as an example, with reference to fig. 2 and 4.
Fig. 4 is an information recommendation method applied to an e-commerce platform, which includes the following steps:
step S401: and creating a marketing plan by using a marketing algorithm module according to the recommendation target parameters.
The marketing algorithm module may be as shown in fig. 2, and the recommendation target parameters may be a return on investment, an inventory consumption parameter, a total amount of value, and a number of people to be recommended. The marketing algorithm module may output the selection recommendations, the quota recommendations, and the crowd recommendations. Wherein, the selection recommendation is also commodity information; the quota recommendation is also the specific commodity information privilege quota; the crowd recommendation refers to the crowd recommending commodity information.
And S402, generating commodity recommendation content by using a marketing content generation module, and obtaining marketing commodity aggregated content through aggregation.
As shown in fig. 2, while creating the marketing plan, the commodity information output by the marketing algorithm module may be used as an input of the marketing content generation module, and recommended content related to the commodity may be obtained through NLP (Natural Language Processing) and image Processing technology, including a recommended short title, a recommended short text, recommended document information, a recommended short video, a comment solarization slip, and the like.
When the commodity information is a commodity category, commodities with historical order numbers larger than a preset threshold value in the commodity category can be used as the input of the marketing content generation module; the commodities with the commodity sales volume larger than the preset threshold value in the historical orders can also be used as the input of the marketing content generation module; and commodities with the historical order transaction total amount larger than a preset threshold value can be used as the input of the marketing content generation module.
And aggregating the generated recommended content and the marketing commodity list to obtain the marketing commodity aggregated content. The marketing commodity aggregated content can comprise multi-dimensional information such as marketing commodity information, preferential information, sun lists, comments, marketing copy and the like.
And S403, storing the marketing commodity aggregated content into a database.
As shown in fig. 2, the marketing aggregate content may be stored in a database through a data storage module, which may be a MYSQL or Redis database, or an hbsase or Elasticsearch database.
And S404, receiving the trigger of the user aiming at the target area, and sequencing the plurality of commodities to be recommended corresponding to the target area by utilizing a sequencing model in the recommending module.
Where the target area may be a content portal, such as a find good-recommend, find good-public-tombstone, shop-and-go, bonus-order, etc. portal. Or marketing entries such as item detail coupons, shopping cart coupons, advertising impressions, and kiosks.
The sorting result of the sorting module is also the recommended sequence of the multiple commodities to be recommended for the current user.
And S405, recommending the plurality of commodities to be recommended and the aggregated contents of the marketing commodities corresponding to the commodities to be recommended in a display page according to the result of the sequencing model.
As shown in fig. 2, in the e-commerce platform, the aggregated content of the marketing goods stored in the database is displayed and recommended in the display page according to the ranking result of the goods to be recommended through an RPC or HTTP protocol.
By the information recommendation method, the user can browse the commodity recommendation content interested by the user through the triggering of the marketing entry by the user, the click conversion rate of the commodity can be improved, and the marketing efficiency is further improved.
Likewise, the user may be caused to browse to the marketing aggregate information by the user's triggering of the content portal. For example, when a user triggers a discovery entrance, a to-be-recommended commodity corresponding to the discovery entrance and a plurality of marketing aggregate information corresponding to the commodity can be recommended to the user, so that the user can quickly find preferential information corresponding to a plurality of marketing related to the interested commodity, the operation of searching the preferential information by the user is reduced, and the user experience is improved.
According to the information recommendation method provided by the embodiment of the invention, the aggregated content of the articles corresponding to the target area can be recommended to the user through the display page according to the sequencing result of the sequencing model on the information of the articles corresponding to the target area, so that the user can obtain the interested content recommendation without additional operation, the effectiveness and efficiency of information recommendation are improved, and the user experience is further improved.
Furthermore, the information of a plurality of articles can be used as the input of a content generation model, the recommended content can be generated and aggregated for the information of the plurality of articles, and the workload of information recommendation is reduced; meanwhile, the sorting model is optimized according to feedback information recommended by the user aiming at the information, so that the recommendation accuracy of the sorting model is improved, and the click conversion rate of the articles is improved.
As shown in fig. 5, an embodiment of the present invention provides an information recommendation apparatus 500, including: an acquisition module 501, a sorting module 502 and a recommendation module 503; wherein the content of the first and second substances,
the obtaining module 501 is configured to, in response to a trigger of a user for a target area in a display page, obtain a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information;
the sorting module 502 is configured to take the first user characteristic and the plurality of item information as inputs of a sorting model, and determine a sorting result of the plurality of item information according to an output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively;
the recommending module 503 is configured to determine target information to be recommended according to the sorting result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information; and recommending the target information through the display page.
In an embodiment of the present invention, with continued reference to fig. 5, the information recommendation apparatus may further include: an aggregation module 504; wherein the content of the first and second substances,
the aggregation module 504 is configured to use the plurality of item information as an input of a content generation model, so as to obtain the recommended content corresponding to each of the plurality of item information; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information; and aggregating the recommended contents to obtain the item aggregated contents.
In an embodiment of the present invention, the obtaining module 501 is configured to obtain, before receiving a trigger of a user for a target area in the presentation page, a recommendation target parameter, where the recommendation target parameter includes any one or more of the following items: return on investment, inventory consumption parameters, total value and the number of people to be recommended; and determining a plurality of item information corresponding to the target area according to the recommended target parameter, the type of the target area and the item information which can be displayed on the display page.
In an embodiment of the present invention, the obtaining module 501 is configured to determine that the item information includes an item category; determining target article information from the article information corresponding to the article category; the target item information is item information of which the historical order quantity, the order completion quantity and/or the order value are/is larger than a preset threshold value; and taking the target item information as an input of the content generation model.
In an embodiment of the present invention, the sorting module 502 is configured to obtain feedback information of the user for the target information through the display page; and optimizing the sequencing model according to the feedback information and preset time length.
In an embodiment of the present invention, the ranking module 502 is configured to determine that the ranking model is trained based on an LR algorithm, a GBDT algorithm, an Xgboost algorithm, a LightGBM algorithm, an xDeepFM algorithm, a deep fm algorithm, and an AutoInt algorithm.
In an embodiment of the present invention, the recommending module 503 is configured to, before determining target information to be recommended according to a sorting result and item aggregation content, use the multiple items of information as an input of a content generating model, and obtain the recommended content corresponding to each of the multiple items of information; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information; and aggregating the recommended contents to obtain the item aggregated contents.
According to the information recommendation device provided by the embodiment of the invention, the item aggregation content corresponding to the target area can be recommended to the user through the display page according to the sequencing result of the sequencing model on the item information corresponding to the target area. Therefore, the operation times of the user are reduced, the interested content can be accurately recommended to the user, the effectiveness and efficiency of information recommendation are improved, and the user experience is further improved.
Furthermore, the information of a plurality of articles can be used as the input of the content generation model, the recommended content can be generated and aggregated for the information of the plurality of articles, the data volume of the recommended information is reduced, and the user can quickly acquire the recommended key points from the recommended information, so that the accuracy of the recommended information is improved. In addition, the sorting model is optimized according to the feedback information recommended by the user aiming at the information, so that the recommendation accuracy of the sorting model is improved, and the click conversion rate of the articles is improved.
Fig. 6 shows an exemplary system architecture 600 to which the information recommendation method or the information recommendation apparatus according to the embodiment of the present invention can be applied.
As shown in fig. 6, the system architecture 600 may include terminal devices 601, 602, 603, a network 604, and a server 605. The network 604 serves to provide a medium for communication links between the terminal devices 601, 602, 603 and the server 605. Network 604 may include various types of connections, such as wire, wireless communication links, or fiber optic cables, to name a few.
A user may use the terminal devices 601, 602, 603 to interact with the server 605 via the network 604 to receive or send messages or the like.
The terminal devices 601, 602, 603 may be various electronic devices having a display screen and supporting information browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 605 may be a server that provides various services, such as a server that provides information recommendation for a user to trigger a presentation page target area using the terminal devices 601, 602, 603. The information recommendation server may sort a plurality of item information corresponding to the target area, generate item aggregation content, determine target information to be recommended, and recommend the target information on the display page to be displayed in the terminal devices 601, 602, and 603.
It should be understood that the number of terminal devices, networks, and servers in fig. 6 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
Referring now to FIG. 7, shown is a block diagram of a computer system 700 suitable for use with a terminal device implementing an embodiment of the present invention. The terminal device shown in fig. 7 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 7, the computer system 700 includes a Central Processing Unit (CPU)701, which can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)702 or a program loaded from a storage section 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data necessary for the operation of the system 700 are also stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
The following components are connected to the I/O interface 705: an input portion 706 including a keyboard, a mouse, and the like; an output section 707 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 708 including a hard disk and the like; and a communication section 709 including a network interface card such as a LAN card, a modem, or the like. The communication section 709 performs communication processing via a network such as the internet. A drive 710 is also connected to the I/O interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that the computer program read out therefrom is mounted in the storage section 708 as necessary.
In particular, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 709, and/or installed from the removable medium 711. The computer program performs the above-described functions defined in the system of the present invention when executed by the Central Processing Unit (CPU) 701.
It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules described in the embodiments of the present invention may be implemented by software or hardware. The described modules may also be provided in a processor, which may be described as: a processor comprises an obtaining module, a sorting module and a recommending module. The names of the modules do not limit the modules themselves in some cases, and for example, the acquisition module may be further described as a "module for acquiring article information".
As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to comprise: responding to the trigger of a user for a target area in a display page, and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information; taking the first user characteristic and the plurality of item information as input of a sorting model, and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively; determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information; and recommending the target information through the display page.
According to the technical scheme of the embodiment of the invention, the item aggregation content corresponding to the target area can be recommended to the user through the display page according to the sequencing result of the sequencing model on the item information corresponding to the target area. Therefore, the operation times of the user are reduced, the interested content can be accurately recommended to the user, the effectiveness and efficiency of information recommendation are improved, and the user experience is further improved.
Furthermore, the information of a plurality of articles can be used as the input of the content generation model, the recommended content can be generated and aggregated for the information of the plurality of articles, the data volume of the recommended information is reduced, and the user can quickly acquire the recommended key points from the recommended information, so that the accuracy of the recommended information is improved. In addition, the sorting model is optimized according to the feedback information recommended by the user aiming at the information, so that the recommendation accuracy of the sorting model is improved, and the click conversion rate of the articles is improved.
The above-described embodiments should not be construed as limiting the scope of the invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions can occur, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. An information recommendation method, comprising:
responding to the trigger of a user for a target area in a display page, and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information;
taking the first user characteristic and the plurality of item information as input of a sorting model, and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively;
determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information;
and recommending the target information through the display page.
2. The method according to claim 1, before determining the target information to be recommended according to the sorting result and the item aggregation content, further comprising:
the plurality of item information is used as the input of a content generation model to obtain the recommended content corresponding to the plurality of item information respectively; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information;
and aggregating the recommended contents to obtain the item aggregated contents.
3. The method of claim 2, wherein before receiving the trigger for the target area by the user in the presentation page, further comprising:
acquiring recommendation target parameters, wherein the recommendation target parameters comprise any one or more of the following items: return on investment, inventory consumption parameters, total value and the number of people to be recommended;
and determining a plurality of item information corresponding to the target area according to the recommended target parameter, the type of the target area and the item information which can be displayed on the display page.
4. The method of claim 3, wherein the item information includes an item category;
determining target article information from the article information corresponding to the article category; the target item information is item information of which the historical order quantity, the order completion quantity and/or the order value are/is larger than a preset threshold value;
and taking the target item information as an input of the content generation model.
5. The method of claim 1, further comprising:
feedback information of the user aiming at the target information is obtained through the display page;
and optimizing the sequencing model according to the feedback information and preset time length.
6. The method of claim 1,
the sequencing model is obtained based on LR algorithm, GBDT algorithm, Xgboost algorithm, LightGBM algorithm, xDeeepFM algorithm, DeepFM algorithm and AutoInt algorithm training.
7. An information recommendation apparatus, comprising: the system comprises an acquisition module, a sorting module and a recommendation module; wherein the content of the first and second substances,
the acquisition module is used for responding to the trigger of a user aiming at a target area in a display page and acquiring a first user characteristic corresponding to the user; the target area corresponds to a plurality of item information;
the sorting module is used for taking the first user characteristic and the plurality of item information as input of a sorting model and determining a sorting result of the plurality of item information according to the output of the sorting model; the ranking model is obtained by training according to a plurality of second user characteristics and historical behavior information corresponding to the second user characteristics respectively;
the recommending module is used for determining target information to be recommended according to the sequencing result and the item aggregation content; the item aggregation content is obtained by aggregation according to the recommended content of the plurality of item information; and recommending the target information through the display page.
8. The apparatus of claim 7, further comprising: a polymerization module; wherein the content of the first and second substances,
the aggregation module is configured to use the plurality of item information as an input of a content generation model to obtain the recommended content corresponding to each of the plurality of item information; the recommended content includes one or more of: recommending short titles, recommending short texts, recommending file information, comment information and recommending short video information; and aggregating the recommended contents to obtain the item aggregated contents.
9. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-6.
10. A computer-readable medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-6.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210092529.5A CN114493786A (en) | 2022-01-26 | 2022-01-26 | Information recommendation method and device |
| PCT/CN2022/124280 WO2023142520A1 (en) | 2022-01-26 | 2022-10-10 | Information recommendation method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210092529.5A CN114493786A (en) | 2022-01-26 | 2022-01-26 | Information recommendation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114493786A true CN114493786A (en) | 2022-05-13 |
Family
ID=81475157
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210092529.5A Pending CN114493786A (en) | 2022-01-26 | 2022-01-26 | Information recommendation method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN114493786A (en) |
| WO (1) | WO2023142520A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023142520A1 (en) * | 2022-01-26 | 2023-08-03 | 北京沃东天骏信息技术有限公司 | Information recommendation method and apparatus |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117151828B (en) * | 2023-10-30 | 2024-01-30 | 建信金融科技有限责任公司 | Recommended article pool processing method, device, equipment and medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020194B (en) * | 2018-08-09 | 2021-10-08 | 南京尚网网络科技有限公司 | Resource recommendation method, device and medium |
| CN110633419B (en) * | 2019-09-20 | 2022-10-21 | 百度在线网络技术（北京）有限公司 | Information pushing method and device |
| CN110941740B (en) * | 2019-11-08 | 2023-07-14 | 深圳市雅阅科技有限公司 | Video recommendation method and computer-readable storage medium |
| CN111310050B (en) * | 2020-02-27 | 2023-04-18 | 深圳大学 | Recommendation method based on multilayer attention |
| CN114493786A (en) * | 2022-01-26 | 2022-05-13 | 北京沃东天骏信息技术有限公司 | Information recommendation method and device |
-
2022
- 2022-01-26 CN CN202210092529.5A patent/CN114493786A/en active Pending
- 2022-10-10 WO PCT/CN2022/124280 patent/WO2023142520A1/en unknown
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023142520A1 (en) * | 2022-01-26 | 2023-08-03 | 北京沃东天骏信息技术有限公司 | Information recommendation method and apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023142520A1 (en) | 2023-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2992563C (en) | Method and system for applying probabilistic topic models to content in a tax environment to improve user satisfaction with a question and answer customer support system | |
| CN104281961B (en) | For the advertisement in on-line system and the quality score system of content | |
| US10423999B1 (en) | Performing personalized category-based product sorting | |
| US11243992B2 (en) | System and method for information recommendation | |
| WO2023142520A1 (en) | Information recommendation method and apparatus | |
| CN112115363A (en) | Recommendation method, computing device and storage medium | |
| CN108369709A (en) | Network-based ad data service delay reduces | |
| CN104246755A (en) | Method and system to provide video-based search results | |
| CN108632311A (en) | Information-pushing method and device | |
| JP6698040B2 (en) | Generation device, generation method, and generation program | |
| CN111429214B (en) | Transaction data-based buyer and seller matching method and device | |
| CN113763019A (en) | User information management method and device | |
| CN109840788A (en) | For analyzing the method and device of user behavior data | |
| CN112749323A (en) | Method and device for constructing user portrait | |
| CN113011911B (en) | Data prediction method and device based on artificial intelligence, medium and electronic equipment | |
| CN109075987A (en) | Optimize digital assembly analysis system | |
| CN113495991A (en) | Recommendation method and device | |
| CN116955817A (en) | Content recommendation method, device, electronic equipment and storage medium | |
| CN110766488A (en) | Method and device for automatically determining theme scene | |
| CN113449175A (en) | Hot data recommendation method and device | |
| KR101985603B1 (en) | Recommendation method based on tripartite graph | |
| CN113627454A (en) | Article information clustering method, pushing method and pushing device | |
| CN111460300A (en) | Network content pushing method and device and storage medium | |
| CN113326680A (en) | Method and device for generating table | |
| CN113159877A (en) | Data processing method, device, system and computer readable storage medium |
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 |