CN108960958B - Item recommendation method and device - Google Patents
Item recommendation method and device Download PDFInfo
- Publication number
- CN108960958B CN108960958B CN201810395304.0A CN201810395304A CN108960958B CN 108960958 B CN108960958 B CN 108960958B CN 201810395304 A CN201810395304 A CN 201810395304A CN 108960958 B CN108960958 B CN 108960958B
- Authority
- CN
- China
- Prior art keywords
- recommended
- value
- article
- item
- recommendation
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000012545 processing Methods 0.000 claims abstract description 52
- 238000011156 evaluation Methods 0.000 description 49
- 230000032683 aging Effects 0.000 description 14
- 230000006399 behavior Effects 0.000 description 10
- 230000000694 effects Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 229940037201 oris Drugs 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
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
Abstract
The application discloses an article recommendation method and device. The method comprises the steps of obtaining N target objects and recommendation factors corresponding to each target object according to historical behavior data of a user, wherein N is an integer larger than 1; obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article; and performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article. The method and the device solve the technical problem that articles with high relevance can be recommended to the user in a large number of piled places, and reading fatigue of the user is caused.
Description
Technical Field
The application relates to the technical field of data application, in particular to an article recommendation method and device.
Background
With the rapid development of the internet, the application scenes of the article recommendation method are more and more, and the article recommendation method is widely integrated into a plurality of business application systems, and the well-known Netflix online video recommendation system, Amazon online shopping mall, today's top line and the like. In fact, most information platforms apply item recommendation methods to different degrees, such as news in New wave, today's headlines, etc. The related recommendation provided by the article recommendation method can help a user to better find the articles related to the current articles, related commodities can be sold in a binding mode in an e-commerce platform, related information recommended by the article recommendation method can be used for deep reading or interest discovery and the like in an information platform, and the method is an important application scene for showing the value of a recommendation system.
In the prior art, an article recommendation method is to recommend based on user behaviors, generally based on historical behavior records of users, determine similarity between articles, and recommend according to the similarity of the articles.
In the process of implementing the embodiment of the present application, the inventor finds that the prior art has at least the following technical problems:
in the prior art, articles about a topic may appear in large quantities at a moment, and the correlation among the articles is high, so that the articles with high correlation are recommended to a user in a large quantity, reading fatigue of the user is caused, and the experience effect of the user is influenced.
Disclosure of Invention
The application mainly aims to provide an article recommendation method and device to solve the problem that articles with high relevance can be recommended to a user in a large number of piled places, and reading fatigue of the user is caused.
In order to achieve the above object, in a first aspect, an embodiment of the present application provides an item recommendation method.
The item recommendation method according to the application comprises the following steps:
acquiring N target articles and recommendation factors corresponding to each target article according to the historical behavior data of the user, wherein N is an integer greater than 1;
obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article;
and performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article.
Optionally, performing data processing on the first recommended value corresponding to each recommended item according to the first recommended value ranking and the recommendation factor corresponding to each target item, and obtaining a second recommended value corresponding to each recommended item, including:
determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item;
according to the first recommendation value sequence and the recommendation factor corresponding to each recommended article, acquiring the sequence of each recommendation relative to the corresponding recommendation factor, and determining the sequence number of each recommended article in the corresponding recommendation factor sequence;
and performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence, and acquiring a second recommended value corresponding to each recommended article.
Optionally, when the recommended item has at least two corresponding second recommended values, a maximum second recommended value of the at least two second recommended values of the recommended item is selected as the second recommended value corresponding to the recommended item.
Optionally, the second recommended value of each recommended item is sorted in descending order of the second recommended value of each recommended item, and the item recommendation method further includes:
in the second recommended value sequence of each recommended article, the similarity value of each recommended article relative to the previous recommended article is obtained, wherein the similarity value included in the recommended article ranked at the first position is a first preset value;
and performing data processing on the second recommended value of each recommended item according to the similarity value included by each recommended item, and acquiring a third recommended value of each recommended item, so as to obtain a third recommended value sequence of each recommended item.
Optionally, the item recommendation method further includes:
and acquiring the generation time of each recommended article, performing data processing on the third recommended value of each recommended article according to the current time and the generation time of each recommended article, and acquiring the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
In a second aspect, an article recommendation device provided in an embodiment of the present application includes:
the system comprises a first acquisition module, a second acquisition module and a recommendation module, wherein the first acquisition module is used for acquiring N target articles and recommendation factors corresponding to each target article according to historical behavior data of a user, and N is an integer greater than 1;
the second obtaining module is used for obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article;
and the third acquisition module is used for performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article, and acquiring a second recommended value corresponding to each recommended article, so that the second recommended value sequence of each recommended article is obtained.
Optionally, the third obtaining module is configured to:
determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item;
according to the first recommendation value sequence and the recommendation factor corresponding to each recommended article, acquiring the sequence of each recommendation relative to the corresponding recommendation factor, and determining the sequence number of each recommended article in the corresponding recommendation factor sequence;
and performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence, and acquiring a second recommended value corresponding to each recommended article.
Optionally, the third obtaining module is further configured to, when the recommended article has at least two corresponding second recommended values, select a largest second recommended value of the at least two second recommended values of the recommended article as the second recommended value corresponding to the recommended article.
Optionally, the article recommendation device further includes a fourth obtaining module and a fifth obtaining module; the second recommended value of each recommended item is sorted into the second recommended value of each recommended item in descending order;
the fourth obtaining module is used for obtaining the similarity value of each recommended article relative to the previous recommended article in the second recommended value sequence of each recommended article, wherein the similarity value included in the recommended article arranged at the first position is a first preset value;
and the fifth obtaining module is used for performing data processing on the second recommended value of each recommended article according to the similarity value included by each recommended article, and obtaining the third recommended value of each recommended article, so that the third recommended value sequence of each recommended article is obtained.
Optionally, the article recommendation device further includes a sixth obtaining module;
and the sixth obtaining module is used for obtaining the generation time of each recommended article, performing data processing on the third recommended value of each recommended article according to the current time and the generation time of each recommended article, and obtaining the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
According to the item recommendation method provided by the embodiment of the application, N target items and recommendation factors corresponding to each target item are obtained according to historical behavior data of a user, wherein N is an integer larger than 1; obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article; and performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article. Therefore, the recommended articles with high correlation are processed according to the recommendation factor and the first recommendation value sequence, the purpose that the recommendation values of the recommended articles are calculated by multiple recommendation dimensions is achieved, and the topic diversity of the recommendation result is improved, the diversity of the recommended articles is achieved, the technical effect of the reading experience of a user is improved, and the technical problem that the articles with high correlation can be recommended to the user in a large number of bundles and the reading fatigue of the user is caused is solved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, serve to provide a further understanding of the application and to enable other features, objects, and advantages of the application to be more apparent. The drawings and their description illustrate the embodiments of the invention and do not limit it. In the drawings:
FIG. 1 is a flow chart of a method of item recommendation according to an embodiment of the present application;
FIG. 2 is a flowchart of a step S300 according to an embodiment of the present application;
FIG. 3 is a flow chart of another method of item recommendation according to an embodiment of the present application;
FIG. 4 is a flowchart of a step S500 according to an embodiment of the present application;
FIG. 5 is a flow chart of another method of item recommendation according to an embodiment of the present application;
FIG. 6 is a flowchart of a step S600 according to an embodiment of the present application;
FIG. 7 is a flowchart of a step S620 according to an embodiment of the present application;
FIG. 8 is a flowchart of a step S621 according to an embodiment of the present application;
FIG. 9 is a flowchart of a step S200 according to an embodiment of the present application;
FIG. 10 is a schematic diagram of an article recommendation device according to an embodiment of the present application;
FIG. 11 is a schematic diagram of another article recommendation device according to an embodiment of the present application;
fig. 12 is a schematic structural diagram of another article recommendation device according to an embodiment of the application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
An embodiment of the present application provides an item recommendation method, as shown in fig. 1, the method includes the following steps S100 to S300:
s100, obtaining N target articles and recommendation factors corresponding to each target article according to the historical behavior data of the user, wherein N is an integer larger than 1.
In this embodiment, the target object may also be information in an information platform, for example, the target object may be a sports news or an entertainment news; the target object may also be a video in a video platform, for example, the target object may be a movie or a tv show; the target object can be a commodity or a service item in an e-commerce platform, for example, the target object can be an electronic product or a consultation service item on a certain e-commerce platform. Correspondingly, the recommended article related to the target article can be information in an information platform; the video platform can also be used for video, and goods or service items in the E-commerce platform can also be used.
In implementation, N target items are obtained according to the user historical behavior data, specifically: extracting M operation articles from the latest operation records (operation articles on the current day) in the user history browsing according to the category information, the relevance and the operation time of the operation articles as target articles of recent interest; by using historical browsing data of a user (generally, 3-5 days of operation articles are adopted), according to the category information and the correlation of the operation articles, Q operation articles extracted from the operation articles are taken as target articles of historical interest, wherein M and Q are integers which are greater than or equal to 0, and M + Q is equal to N. And generates recommendation factors for each target item, wherein each recommendation factor can be a topic (such as a topic related to a certain entertainment star) or business information or experience information. The service information can be set for service requirements, for example, articles related to certain topics need to be popularized; the experience information may be historical experience, for example, when the title of an item contains a specific word, for example, when the target item is sports news, the content of the target item contains "mei xi", and the recommendation factor is related to "mei xi".
Wherein, the value of N can be 2, 3, 4, 5, 6 … …
S200, obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article.
In an implementation, the method for obtaining a plurality of recommended items related to the target item may be a plurality of recommended items related to the target item generated based on a content recommendation method, a plurality of recommended items related to the target item generated according to a user collaborative recommendation method or an item collaborative recommendation method, a plurality of recommended items related to the target item generated according to a business rule recommendation method, or a union set generated by the methods, where the union set includes the plurality of recommended items related to the target item generated by the methods, and the method obtains the plurality of recommended items related to the target item and obtains a recommendation value corresponding to each recommended item.
S300, performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article, and acquiring a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article.
In this embodiment, each recommended item corresponds to at least one target item, and each target item corresponds to one recommendation factor, so that each recommended item corresponds to at least one recommendation factor, and the first recommendation value of the recommended item is calculated according to the serial number position of the recommended item in the first recommendation value ranking and the corresponding recommendation factor, so as to obtain the first recommendation value of the recommended item. And then obtaining a second recommended value sequence of each recommended item. In this way, the recommended articles with high relevance are processed according to the recommendation factor and the first recommendation value sequence, the purpose that the recommendation values of the recommended articles are calculated by multiple recommendation dimensions is achieved, and the topic diversity of the recommendation result is improved, so that the diversity of the recommended articles is realized, and the technical effect of the reading experience of a user is improved.
As shown in fig. 2, optionally, in S300, performing data processing on the first recommended value corresponding to each recommended item according to the first recommended value ranking and the recommendation factor corresponding to each target item, and obtaining the second recommended value ranking corresponding to each recommended item, so as to obtain the second recommended value ranking of each recommended item, including the following steps S310 to S330:
s310, determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item.
In this embodiment, each recommended item corresponds to at least one target item, and each target item corresponds to one recommendation factor, so that each recommended item corresponds to at least one recommendation factor.
S320, acquiring the corresponding recommendation factor sequence of each recommendation object according to the first recommendation value sequence and the recommendation factor corresponding to each recommendation object, and determining the sequence number of each recommendation object in the corresponding recommendation factor sequence.
In this embodiment, according to the sequence number position of the recommended item in the first recommendation value sequence, the sequence between a plurality of recommended items corresponding to the same recommendation factor is determined, so as to determine the sequence number of each recommended item in the corresponding recommendation factor sequence.
For example, there are 3 recommended articles corresponding to one recommendation factor, which are respectively a first recommended article, a second recommended article, and a third recommended article, where sequence numbers of the first recommended article, the second recommended article, and the third recommended article in the first recommended value ranking are respectively 2 nd, 9 th, and 4 th, and then sequence numbers of the 3 recommended articles relative to the corresponding recommendation factor ranking are respectively: the first recommended article corresponds to the serial number 1, the second recommended article corresponds to the serial number 3, and the third recommended article corresponds to the serial number 2.
S330, performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence, and acquiring a second recommended value corresponding to each recommended article, so as to obtain the second recommended value of each recommended article.
In this embodiment, data processing is performed on the first recommended value corresponding to the recommended item according to the serial number of the recommended item in the corresponding recommendation factor ranking, and the data processing may be implemented by the following formula:
in the formula, rec _ score is on the left side of equal signiIs the second recommended value of the ith recommended item, rec _ score on the right side of the equal signiIs the first recommendation value for the ith recommended item,
is the serial number of the ith recommended item in the corresponding recommendation factor ranking (or the occurrence number of the recommendation factors when the first recommendation value ranking is descending), constrsnThe recommendation factor is a preset constant corresponding to each recommendation factor, optionally, the preset constant may select different values according to different recommendation factors, and the preset constant may also adopt a preset fixed value, for example, the value may be 1.
In addition, optionally, when the recommended item has at least two corresponding second recommended values, the largest second recommended value of the at least two second recommended values of the recommended item is selected as the second recommended value corresponding to the recommended item.
As shown in fig. 3, optionally, the second recommended value ranking of each recommended item is a descending order of the second recommended value of each recommended item, and the item recommendation method further includes the following steps S400 and S500:
s400, in the second recommended value sequence of each recommended article, the similarity value of each recommended article relative to the previous recommended article is obtained, wherein the similarity value included in the recommended article arranged at the first position is a first preset value.
In this embodiment, the recommended item ranked first in the descending order of the second recommended value may have no recommended item in front of it, and therefore, the similarity value of the recommended item ranked first may be set to a first preset value, where optionally, the value of the first preset value may be 0.
In implementation, when the similarity value of each recommended item with respect to the previous recommended item is obtained, the similarity value of the recommended item with respect to the previous recommended item may be calculated by using titles (titles) included in two adjacent recommended items, where the calculation method may adopt an existing similarity calculation method.
And S500, performing data processing on the second recommended value of each recommended article according to the similarity value included by each recommended article, and acquiring a third recommended value of each recommended article, so as to obtain a third recommended value sequence of each recommended article.
In this embodiment, the second recommendation value corresponding to each recommended item is subjected to data processing according to the similarity value included in each recommended item, so that the second recommendation value corresponding to each recommended item is adjusted according to the content information, and the first recommendation value of each recommended item is generated. Therefore, the content similarity of two adjacent recommended articles can be prevented, and the recommendation experience of the user is further influenced.
As shown in fig. 4, optionally, in S500, performing data processing on the second recommended value of each recommended item according to the similarity value included in each recommended item, and obtaining a third recommended value of each recommended item, so as to obtain a third recommended value ranking of each recommended item, including the following steps S510 and S520:
and S510, acquiring a dissimilarity value of each recommended item relative to the previous recommended item according to the similarity value included by each recommended item.
And S520, performing data processing on the second recommended value corresponding to each recommended article according to the dissimilarity value included in each recommended article, and acquiring a third recommended value of each recommended article, so as to obtain a third recommended value sequence of each recommended article.
In this embodiment, the third recommended value of the recommended item may be a product of a dissimilarity value of the recommended item with respect to a previous recommended item and the second recommended value corresponding to the recommended item.
For example, if the second recommended value corresponding to a recommended item is 0.9 and the dissimilarity value of the recommended item with respect to the previous recommended item is 0.3, then the third recommended value of the recommended item is 0.27 (the product of 0.9 and 0.3).
Optionally, when the dissimilarity value of the recommended item relative to the previous recommended item is smaller than a second preset value, the second preset value is selected as the dissimilarity of the recommended item.
In implementation, the second preset value may be set to prevent that the dissimilarity value of the recommended item with respect to the previous recommended item is too small, which may result in that the third recommended value of the recommended item is too small, and further affect the processing operation of other subsequent steps on the recommended item.
In this embodiment, data processing is performed on the second recommended value corresponding to each recommended item according to the similarity value included in each recommended item, so as to obtain a third recommended value of each recommended item, where the data processing may be implemented by the following formula:
rec_scorei＝rec_scorei*max((1-sim(itemi,itemi-1)),a)
wherein, the left rec _ score of equal signiIs the third recommended value of the ith recommended item, rec _ score on the right side of the equal signiIs the second recommended value, sim (item), for the ith recommended itemi,itemi-1) (ii) is the similarity value of the ith recommended item to the previous recommended item, (1-sim (item)i,itemi-1) And a is a second preset value, where the value of the second preset value may be 0.08, 0.1, 0.2, and the like, and those skilled in the art may specifically set the value according to actual needs.
As shown in fig. 5, optionally, the item recommendation method further includes the following step S600:
s600, obtaining the generation time of each recommended article, performing data processing on the third recommended value of each recommended article according to the current time and the generation time of each recommended article, and obtaining the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
In this embodiment, the generation time of the recommended item may be a Uniform Resource Locator (URL) generation time of the recommended item; the current time is the time at which the recommendation was generated.
In the implementation, the third recommended value of each recommended item is subjected to data processing according to the current time and the generation time of each recommended item, so that in the process of performing data processing on the third recommended value of each recommended item, the influence of a time factor on the third recommended value is added, the timeliness of the recommended items is considered, the purpose of recommending the recommended items with good timeliness to the user is achieved, the recommended items with good timeliness and high relevance to the interest of the user are recommended to the user, and the technical effect of improving the interest of the user on the recommended items is achieved.
As shown in fig. 6, optionally, in S600, performing data processing on the third recommended value of each recommended item according to the current time and the generation time of each recommended item, and obtaining a fourth recommended value of each recommended item, so as to obtain a fourth recommended value ranking of each recommended item, including the following steps S610 to S630:
s610, obtaining the generation duration of each recommended item according to the current time and the generation time of each recommended item.
In implementation, the generation time of the recommended item, that is, the length of time that the recommended item exists from the generation to the current time, may be obtained by subtracting the generation time of the recommended item from the current time, and the time unit of the generation time may be seconds, minutes, hours, days, and the like.
For example, if a recommended item is a piece of sports news about the west, the generation time of the sports news about the west is 18 hours, 35 minutes and 18 seconds at 4 months, 17 days, 2018 years, 4 months, 19 days, 19 hours, 45 minutes and 30 seconds at 2018 years, the generation time of the sports news about the west is specifically 2 days, 1 hour, 10 minutes and 12 seconds, the generation time of the sports news about the west is 49.17 hours when the time unit is hour, and the generation time of the sports news about the west is 177012 seconds when the time unit is second.
S620, acquiring the aging factor of each recommended item according to the first preset time duration evaluation value and the generation time duration of each recommended item.
In this embodiment, the first preset time duration evaluation value is a preset value, and may be set or adjusted according to actual needs, and the first preset time duration evaluation value may be a constant, or may be a value having a time unit.
In implementation, the aging factor of the recommended article is obtained by calculating the production time length of the recommended article and the first preset time length evaluation value.
For example, if a recommended item is social news about an ecological environment, the generation time of the social news about the ecological environment is 30 minutes, the first preset time duration evaluation value is 1440 minutes, and the operation is a ratio of the first preset time duration to the generation time duration, then the aging factor of the social news about the ecological environment is 48 (a ratio of 1440 minutes to 30 minutes).
S630, calculating the third recommended value of each recommended article according to the aging factor of each recommended article, and obtaining the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
In this embodiment, the third recommended value of the recommended item is calculated according to the aging factor of the recommended item, so as to obtain a fourth recommended value of the recommended item.
For example, if the operation is the product of the aging factor of the recommended item and the third recommended value of the recommended item, the aging factor of the recommended item is 2, the third recommended value of the recommended item is 0.9001, and then the fourth recommended value of the recommended item is 1.8002(2 is multiplied by 0.9001).
As shown in fig. 7, optionally, the step S620 of obtaining the aging factor of each recommended item according to the first preset time duration evaluation value and the generation time duration of each recommended item includes the following steps S621 to S622:
and S621, acquiring a first time length evaluation value of each recommended item according to the generation time length of each recommended item.
In this embodiment, the evaluation value corresponding to the generation time length may be obtained according to the generation time length of the recommended item, that is, the first time length evaluation value, and the first time length evaluation value may be obtained by performing data processing on the generation time length of the recommended item, for example, by dividing the generation time length of the recommended item by one unit time, so as to obtain the first time length evaluation value corresponding to the generation time length.
For example, if the generation time of a certain recommended item is 4 days and the unit time is day, the first time evaluation value may be 4.
And S622, according to the ratio of the first preset time length evaluation value to the first time length evaluation value of each recommended article, the aging factor of each recommended article.
In this embodiment, the aging factor of the recommended item may be a ratio obtained from a ratio of the first preset time period evaluation value to the first time period evaluation value of the recommended item.
For example, if the first preset time period evaluation value is 1000 and the first time period evaluation value of a recommended item is 5, the aging factor of the recommended item is 200(1000 divided by 5).
As shown in fig. 8, optionally, in step S621, the step of obtaining the first time length evaluation value of each recommended item according to the generation time length of each recommended item includes the following steps S6211 to S6212:
s6211, according to the ratio of the generation time length of each recommended article to the unit time, a second time length evaluation value of each recommended article is obtained.
In this embodiment, the second time period evaluation value of the recommended item may be a ratio obtained from a ratio of the generation time period of the recommended item to the unit time.
For example, if the generation time period of a certain recommended item is 30 hours and the unit time is hours, then the second time period evaluation value of the recommended item is 30, and for example, if the generation time period of another recommended item is 0, that is, if the recommended item is generated at the current time, then the second time period evaluation value of the recommended item is 0.
And S6212, acquiring a first time length evaluation value of each recommended article according to the sum of the second preset time length evaluation value and the second time length evaluation value of each recommended article.
In the present embodiment, to prevent the first time length evaluation value from being 0, the first time length evaluation value is the sum of a second preset time length evaluation value, which is a value greater than 0, and a second time length evaluation value of each recommended item.
For example, the second preset time period evaluation value may be 1, and when the second time period evaluation value of a certain recommended item is 0, the first time period evaluation value of the recommended item is 1.
In this embodiment, in S600, the generation time of each recommended item is obtained, data processing is performed on the third recommended value of each recommended item according to the current time and the generation time of each recommended item, and the fourth recommended value of each recommended item is obtained, specifically:
and adjusting the timeliness of the third recommended value of the recommended article according to the generation time and the current time of the recommended article to obtain a fourth recommended value of the recommended article, wherein the formula is as follows:
wherein, the left rec _ score of equal signiIs the fourth recommended value of the ith recommended item, the rec _ score _ i on the right side of the equal sign is the third recommended value of the ith recommended item, cur _ time is the current time, the time unit is second, and create _ timeiIs the generation time of the ith recommended item, the time unit is second, consttime1The evaluation value of the first preset duration evaluation value may be specifically set by a person skilled in the art according to actual needs, for example, the evaluation value of the first preset duration evaluation value may be 1, 5, 10, 30, 100, 300, 1000, and the like; in the formula, consttime2Is the secondA value of the second preset duration evaluation value may be specifically set by a person skilled in the art according to actual needs, for example, the second preset duration evaluation value may be 1, and the like; in the formula, (cur _ time-create _ timei) Is the generation duration of the ith recommended item, (cur _ time-create _ time)i) /86400 is the second time period evaluation value of the ith recommended item, (consttime2+(cur_time-create_timei) 86400) is the first time length evaluation value of the ith recommended item.
As shown in fig. 9, optionally, step S200, obtaining a plurality of recommended items and a first recommended value corresponding to each recommended item according to the N target items, and obtaining a first recommended value ranking of each recommended item, includes the following steps S210 to S270:
s210, at least one recommendation set is obtained according to each target item, wherein each recommendation set comprises a plurality of recommendation items related to the target item, and each recommendation item corresponds to a fifth recommendation value.
In this embodiment, each target item may be subjected to recommended item screening by using each existing item recommendation method, so as to obtain one recommendation set.
S220, normalization processing is carried out on the fifth recommended values corresponding to the recommended articles in each recommended set, and a sixth recommended value corresponding to each recommended article is obtained.
In this embodiment, normalization processing is performed on the fifth recommended value corresponding to the recommended item included in each recommended set, and a sixth recommended value corresponding to each recommended item is obtained, which may be implemented by the following formula:
wherein, the left rec _ score of equal signiIs the sixth recommended value of the ith recommended item in a certain recommended set, rec _ score on the left side of equal signiIs the fifth recommendation value, rec _ score, for the ith recommended item in a recommendation set1Is the fifth recommended value with the largest value in a certain recommended set.After each recommendation set is normalized in this way, each sixth recommendation value is [0, 1 ]]Within the interval (c).
And S230, fusing all recommendation sets into a recommendation union set, wherein when the same recommended item exists in different recommendation sets, the sixth recommendation value of the recommended item in the recommendation union set is the sum of the sixth recommendation values corresponding to the recommended item in different recommendation sets.
And S240, determining the descending order of the sixth recommended value of each recommended item in the recommended union set.
And S250, acquiring the belonged category of each recommended article.
And S260, obtaining the category sequence of each recommended article according to the descending sequence of the sixth recommended value and the category to which each recommended article belongs, and determining the sequence number of each recommended article in the corresponding category sequence to which each recommended article belongs.
And S270, performing data processing on the sixth recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding belonged category sequence, and acquiring the first recommended value corresponding to each recommended article, so as to obtain the first recommended value sequence of each recommended article.
In this embodiment, data processing is performed on the sixth recommended value corresponding to each recommended item according to the serial number of each recommended item in the corresponding category ranking to which each recommended item belongs, so as to obtain the first recommended value corresponding to each recommended item, which can be implemented according to the following formula:
wherein, the left rec _ score of equal signiIs the first recommended value of the ith recommended item, rec _ score on the right side of the equal signiIs the sixth recommended value for the ith recommended item,
is the serial number (or
Is the number of occurrences of the ith recommended item in the sixth recommended value descending order), constcateIs a class constant that may be set to 1 or other value greater than 0.
According to the item recommendation method provided by the embodiment of the application, N target items and recommendation factors corresponding to each target item are obtained according to historical behavior data of a user, wherein N is an integer larger than 1; obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article; and performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article. Therefore, the recommended articles with high correlation are processed according to the recommendation factor and the first recommendation value sequence, the purpose that the recommendation values of the recommended articles are calculated by multiple recommendation dimensions is achieved, and the topic diversity of the recommendation result is improved, the diversity of the recommended articles is achieved, the technical effect of the reading experience of a user is improved, and the technical problem that the articles with high correlation can be recommended to the user in a large number of bundles and the reading fatigue of the user is caused is solved.
Based on the same technical concept as the item recommendation method, an embodiment of the present application further provides an item recommendation device, as shown in fig. 10, including:
the first obtaining module 10 is configured to obtain N target items and recommendation factors corresponding to each target item according to user historical behavior data, where N is an integer greater than 1;
a second obtaining module 20, configured to obtain, according to the N target items, a plurality of recommended items and a first recommended value corresponding to each recommended item, and obtain a first recommended value ranking of each recommended item, where each recommended item is related to at least one target item;
the third obtaining module 30 performs data processing on the first recommended value corresponding to each recommended item according to the first recommended value ranking and the recommendation factor corresponding to each target item, and obtains a second recommended value corresponding to each recommended item, so as to obtain a second recommended value ranking of each recommended item.
Optionally, the third obtaining module 30 is configured to:
determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item;
according to the first recommendation value sequence and the recommendation factor corresponding to each recommended article, acquiring the sequence of each recommendation relative to the corresponding recommendation factor, and determining the sequence number of each recommended article in the corresponding recommendation factor sequence;
and performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence, and acquiring a second recommended value corresponding to each recommended article.
Optionally, the third obtaining module 30 is further configured to, when the recommended article has at least two corresponding second recommended values, select a maximum second recommended value of the at least two second recommended values of the recommended article as the second recommended value corresponding to the recommended article.
As shown in fig. 11, optionally, the item recommendation apparatus further includes a fourth obtaining module 40 and a fifth obtaining module 50; the second recommended value of each recommended item is sorted into the second recommended value of each recommended item in descending order;
a fourth obtaining module 40, configured to obtain, in the second recommended value ranking of each recommended item, a similarity value of each recommended item with respect to a previous recommended item, where a similarity value included in a recommended item ranked first is a first preset value;
a fifth obtaining module 50, configured to perform data processing on the second recommended value of each recommended item according to the similarity value included in each recommended item, and obtain a third recommended value of each recommended item, so as to obtain a third recommended value ranking of each recommended item.
Optionally, the fifth obtaining module 50 is configured to:
acquiring a dissimilarity value of each recommended article relative to the recommended article at the previous position according to the similarity value included by each recommended article;
performing data processing on the second recommended value corresponding to each recommended article according to the dissimilarity value included in each recommended article to obtain a third recommended value of each recommended article, so as to obtain a third recommended value sequence of each recommended article;
optionally, when the dissimilarity value of the recommended item relative to the previous recommended item is smaller than a second preset value, the second preset value is selected as the dissimilarity of the recommended item.
As shown in fig. 12, optionally, the item recommendation apparatus further includes a sixth obtaining module 60;
a sixth obtaining module 60, configured to obtain the generation time of each recommended item, perform data processing on the third recommended value of each recommended item according to the current time and the generation time of each recommended item, and obtain the fourth recommended value of each recommended item, so as to obtain a fourth recommended value ranking of each recommended item.
Optionally, the sixth obtaining module 60 is configured to:
acquiring the generation duration of each recommended item according to the current time and the generation time of each recommended item;
acquiring an aging coefficient of each recommended article according to the first preset time duration evaluation value and the generation time duration of each recommended article;
and calculating the third recommended value of each recommended article according to the aging coefficient of each recommended article to obtain the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
Optionally, the sixth obtaining module 60 is configured to:
acquiring a first time length evaluation value of each recommended item according to the generation time length of each recommended item;
and according to the ratio of the first preset time length evaluation value to the first time length evaluation value of each recommended article, the aging factor of each recommended article.
Optionally, the sixth obtaining module 60 is configured to:
acquiring a second time length evaluation value of each recommended article according to the ratio of the generation time length of each recommended article to the unit time;
and acquiring a first time length evaluation value of each recommended article according to the sum of the second preset time length evaluation value and the second time length evaluation value of each recommended article.
Optionally, the second obtaining module 20 is configured to:
obtaining at least one recommendation set according to each target item, wherein each recommendation set comprises a plurality of recommendation items related to the target item, and each recommendation item corresponds to a fifth recommendation value;
normalizing the fifth recommended value corresponding to the recommended article in each recommended set to obtain a sixth recommended value corresponding to each recommended article;
all recommendation sets are fused into a recommendation union set, wherein when the same recommended articles exist in different recommendation sets, the sixth recommendation value of the recommended article in the recommendation union set is the sum of the sixth recommendation values corresponding to the recommended article in different recommendation sets;
determining a sixth recommended value descending order of each recommended item in the recommended union set;
acquiring the category of each recommended article;
according to the descending order of the sixth recommended value and the category of each recommended article, obtaining the category order of each recommended article, and determining the sequence number of each recommended article in the corresponding category order;
and performing data processing on the sixth recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding belonged category sequence, and acquiring the first recommended value corresponding to each recommended article, so as to obtain the first recommended value sequence of each recommended article.
The article recommendation device provided in the embodiment of the application is configured to, through the first obtaining module 10, obtain N target articles and a recommendation factor corresponding to each target article according to user historical behavior data, where N is an integer greater than 1; a second obtaining module 20, configured to obtain, according to the N target items, a plurality of recommended items and a first recommended value corresponding to each recommended item, and obtain a first recommended value ranking of each recommended item, where each recommended item is related to at least one target item; the third obtaining module 30 is configured to perform data processing on the first recommended value corresponding to each recommended item according to the first recommended value ranking and the recommendation factor corresponding to each target item, and obtain a second recommended value corresponding to each recommended item, so as to obtain a second recommended value ranking of each recommended item. Therefore, the recommended articles with high correlation are processed according to the recommendation factor and the first recommendation value sequence, the purpose that the recommendation values of the recommended articles are calculated by multiple recommendation dimensions is achieved, and the topic diversity of the recommendation result is improved, the diversity of the recommended articles is achieved, the technical effect of the reading experience of a user is improved, and the technical problem that the articles with high correlation can be recommended to the user in a large number of bundles and the reading fatigue of the user is caused is solved.
It will be apparent to those skilled in the art that the modules or steps of the present application described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and they may alternatively be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, or fabricated separately as individual integrated circuit modules, or fabricated as a single integrated circuit module from multiple modules or steps. Thus, the present application is not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (8)
1. An item recommendation method, characterized in that the method comprises:
acquiring N target articles and recommendation factors corresponding to each target article according to the historical behavior data of the user, wherein N is an integer greater than 1;
obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article;
performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so as to obtain a second recommended value sequence of each recommended article;
the second recommended value ranking of each of the recommended items is a descending second recommended value ranking of each of the recommended items, the method further comprising:
in the second recommended value sequence of each recommended article, obtaining the similarity value of each recommended article relative to the previous recommended article, wherein the similarity value included in the recommended article ranked at the first position is a first preset value;
performing data processing on the second recommended value of each recommended item according to the similarity value included in each recommended item, and acquiring a third recommended value of each recommended item, so as to obtain a third recommended value sequence of each recommended item, including:
acquiring a dissimilarity value of each recommended article relative to the recommended article at the previous position according to the similarity value included by each recommended article;
performing data processing on the second recommended value corresponding to each recommended article according to the dissimilarity value included in each recommended article to obtain a third recommended value of each recommended article, so as to obtain a third recommended value sequence of each recommended article;
and when the dissimilarity value of the recommended article relative to the previous recommended article is smaller than a second preset value, selecting the second preset value as the dissimilarity of the recommended article.
2. The item recommendation method according to claim 1, wherein the data processing of the first recommendation value corresponding to each recommended item according to the first recommendation value ranking and the recommendation factor corresponding to each target item to obtain the second recommendation value corresponding to each recommended item comprises:
determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item;
according to the first recommendation value sequence and the recommendation factor corresponding to each recommended article, acquiring the sequence of each recommendation object corresponding to the corresponding recommendation factor, and determining the sequence number of each recommended article in the corresponding recommendation factor sequence;
and performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence to obtain a second recommended value corresponding to each recommended article.
3. The item recommendation method according to claim 2, wherein when at least two corresponding second recommendation values exist for the recommended item, a largest second recommendation value of the at least two second recommendation values for the recommended item is selected as the second recommendation value corresponding to the recommended item.
4. The item recommendation method according to claim 1, further comprising:
and acquiring the generation time of each recommended article, performing data processing on the third recommended value of each recommended article according to the current time and the generation time of each recommended article, and acquiring the fourth recommended value of each recommended article, so as to obtain the fourth recommended value sequence of each recommended article.
5. An item recommendation device, the device comprising:
the system comprises a first acquisition module, a second acquisition module and a recommendation module, wherein the first acquisition module is used for acquiring N target articles and recommendation factors corresponding to each target article according to historical behavior data of a user, and N is an integer greater than 1;
the second obtaining module is used for obtaining a plurality of recommended articles and a first recommended value corresponding to each recommended article according to the N target articles, and obtaining a first recommended value sequence of each recommended article, wherein each recommended article is related to at least one target article;
the third obtaining module is used for performing data processing on the first recommended value corresponding to each recommended article according to the first recommended value sequence and the recommendation factor corresponding to each target article to obtain a second recommended value corresponding to each recommended article, so that the second recommended value sequence of each recommended article is obtained;
the device also comprises a fourth acquisition module and a fifth acquisition module; the second recommended value of each recommended item is sorted into the second recommended value of each recommended item in descending order;
the fourth obtaining module is configured to obtain, in the second recommended value ranking of each recommended item, a similarity value of each recommended item with respect to the previous recommended item, where the similarity value included in the recommended item ranked first is a first preset value;
the fifth obtaining module is configured to perform data processing on the second recommended value of each recommended article according to the similarity value included in each recommended article, and obtain a third recommended value of each recommended article, so as to obtain a third recommended value ranking of each recommended article;
the fifth obtaining module 50 is configured to:
acquiring a dissimilarity value of each recommended article relative to the recommended article at the previous position according to the similarity value included by each recommended article;
performing data processing on the second recommended value corresponding to each recommended article according to the dissimilarity value included in each recommended article to obtain a third recommended value of each recommended article, so as to obtain a third recommended value sequence of each recommended article;
and when the dissimilarity value of the recommended article relative to the previous recommended article is smaller than a second preset value, selecting the second preset value as the dissimilarity of the recommended article.
6. The item recommendation device of claim 5, wherein the third obtaining module is configured to:
determining a recommendation factor corresponding to each recommended item according to each recommended item and the related target item;
according to the first recommendation value sequence and the recommendation factor corresponding to each recommended article, acquiring the sequence of each recommendation object corresponding to the corresponding recommendation factor, and determining the sequence number of each recommended article in the corresponding recommendation factor sequence;
and performing data processing on the first recommended value corresponding to each recommended article according to the serial number of each recommended article in the corresponding recommendation factor sequence to obtain a second recommended value corresponding to each recommended article.
7. The item recommendation device according to claim 6, wherein the third obtaining module is further configured to, when the recommended item has at least two corresponding second recommendation values, select a largest second recommendation value of the at least two second recommendation values of the recommended item as the second recommendation value corresponding to the recommended item.
8. The item recommendation device of claim 5, further comprising a sixth acquisition module;
the sixth obtaining module is configured to obtain a generation time of each recommended item, perform data processing on the third recommended value of each recommended item according to the current time and the generation time of each recommended item, and obtain a fourth recommended value of each recommended item, so as to obtain a fourth recommended value ranking of each recommended item.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810395304.0A CN108960958B (en) | 2018-04-27 | 2018-04-27 | Item recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810395304.0A CN108960958B (en) | 2018-04-27 | 2018-04-27 | Item recommendation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108960958A CN108960958A (en) | 2018-12-07 |
| CN108960958B true CN108960958B (en) | 2022-03-22 |
Family
ID=64498965
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810395304.0A Active CN108960958B (en) | 2018-04-27 | 2018-04-27 | Item recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108960958B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763318B (en) * | 2018-04-27 | 2022-04-19 | 达而观信息科技（上海）有限公司 | Item recommendation method and device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN104598643A (en) * | 2015-02-13 | 2015-05-06 | 成都品果科技有限公司 | Article similarity contribution factor, similarity acquiring method, as well as article recommendation method and system thereof |
| CN104615631A (en) * | 2014-10-29 | 2015-05-13 | 中国建设银行股份有限公司 | Information recommendation method and device |
| CN105183731A (en) * | 2014-06-04 | 2015-12-23 | 腾讯科技（深圳）有限公司 | Method, device, and system for generating recommended information |
| CN105809479A (en) * | 2016-03-07 | 2016-07-27 | 海信集团有限公司 | Item recommending method and device |
| CN106687952A (en) * | 2014-09-26 | 2017-05-17 | 甲骨文国际公司 | Techniques for similarity analysis and data enrichment using knowledge sources |
| CN107092616A (en) * | 2016-11-02 | 2017-08-25 | 北京小度信息科技有限公司 | A kind of object order method and device |
| CN107948748A (en) * | 2017-11-30 | 2018-04-20 | 奇酷互联网络科技（深圳）有限公司 | Recommend method, equipment, mobile terminal and the computer-readable storage medium of video |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100268661A1 (en) * | 2009-04-20 | 2010-10-21 | 4-Tell, Inc | Recommendation Systems |
-
2018
- 2018-04-27 CN CN201810395304.0A patent/CN108960958B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN105183731A (en) * | 2014-06-04 | 2015-12-23 | 腾讯科技（深圳）有限公司 | Method, device, and system for generating recommended information |
| CN106687952A (en) * | 2014-09-26 | 2017-05-17 | 甲骨文国际公司 | Techniques for similarity analysis and data enrichment using knowledge sources |
| CN104615631A (en) * | 2014-10-29 | 2015-05-13 | 中国建设银行股份有限公司 | Information recommendation method and device |
| CN104598643A (en) * | 2015-02-13 | 2015-05-06 | 成都品果科技有限公司 | Article similarity contribution factor, similarity acquiring method, as well as article recommendation method and system thereof |
| CN105809479A (en) * | 2016-03-07 | 2016-07-27 | 海信集团有限公司 | Item recommending method and device |
| CN107092616A (en) * | 2016-11-02 | 2017-08-25 | 北京小度信息科技有限公司 | A kind of object order method and device |
| CN107948748A (en) * | 2017-11-30 | 2018-04-20 | 奇酷互联网络科技（深圳）有限公司 | Recommend method, equipment, mobile terminal and the computer-readable storage medium of video |
Non-Patent Citations (1)
| Title |
|---|
| 基于信任和推荐关系的可信服务发现;刘迎春;《系统工程理论与实践》;20121231;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108960958A (en) | 2018-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108629665B (en) | Personalized commodity recommendation method and system | |
| CN102073699B (en) | For improving the method for Search Results, device and equipment based on user behavior | |
| CN106708817B (en) | Information searching method and device | |
| US20120185481A1 (en) | Method and Apparatus for Executing a Recommendation | |
| US20080215453A1 (en) | Server apparatus, information processing apparatus, and information processing method | |
| CN101446959A (en) | Internet-based news recommendation method and system thereof | |
| CN106878405B (en) | Method and device for adjusting push items | |
| EP1649395A1 (en) | Methods and systems for understanding a meaning of a knowledge item using information associated with the knowledge item | |
| CN105159910A (en) | Information recommendation method and device | |
| US10019419B2 (en) | Method, server, browser, and system for recommending text information | |
| CN110197404B (en) | Personalized long-tail commodity recommendation method and system capable of reducing popularity deviation | |
| CN103559622A (en) | Characteristic-based collaborative filtering recommendation method | |
| CN107423430B (en) | Data processing method, device and computer readable storage medium | |
| CN111310038B (en) | Information recommendation method and device, electronic equipment and computer-readable storage medium | |
| CN108765051B (en) | Item recommendation method and device | |
| CN109241451B (en) | Content combination recommendation method and device and readable storage medium | |
| CN110941762B (en) | Method for improving sequencing result of personalized recommendation system | |
| CN111339416A (en) | Heat recall method and device, electronic equipment and storage medium | |
| CN111915409B (en) | Item recommending method, device, equipment and storage medium based on item | |
| CN112862567A (en) | Exhibit recommendation method and system for online exhibition | |
| CN110309410B (en) | Information recommendation method, platform and computer readable storage medium | |
| CN111310046A (en) | Object recommendation method and device | |
| CN108960958B (en) | Item recommendation method and device | |
| CN108763318B (en) | Item recommendation method and device | |
| CN116823410B (en) | Data processing method, object processing method, recommending method and computing device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: Room 501, 502, 503, No. 66 Boxia Road, China (Shanghai) Pilot Free Trade Zone, Pudong New Area, Shanghai, March 2012 Patentee after: Daguan Data Co.,Ltd. Address before: Room 515, building Y1, No. 112, liangxiu Road, Pudong New Area, Shanghai 201203 Patentee before: DATAGRAND INFORMATION TECHNOLOGY (SHANGHAI) Co.,Ltd. |
|
| CP03 | Change of name, title or address |