CN113158032A - Information pushing method and device - Google Patents

Abstract

The invention discloses an information pushing method and device, and relates to the technical field of computers. One embodiment of the method comprises: screening a target object; selecting anchor point information for a target object, and determining the category of the anchor point information; acquiring behavior data related to anchor point information, and calculating association indexes between the category of the anchor point information and various target categories included in the behavior data; determining a category to be pushed according to the category to which the anchor point information belongs and the associated indexes respectively corresponding to the various target categories; selecting information to be pushed from categories to be pushed according to a preset pushing strategy; and pushing the information to be pushed for the target object when responding to the received page processing request sent by the target object. The embodiment can effectively improve the information pushing accuracy.

Description

Information pushing method and device

Technical Field

The invention relates to the technical field of computers, in particular to an information pushing method and device.

Background

Various applications such as shopping clients, video playing clients, etc. push commodities, videos, advertisements, etc. that are interesting to users for better attracting users.

At present, the information push method mainly analyzes the user's preference based on a large amount of behavior data such as commodities, videos, advertisements and the like frequently browsed or purchased by the user, so as to push the commodities, videos, advertisements and the like which the user is interested in.

In the process of implementing the invention, the inventor finds that at least the following problems exist in the prior art:

for users with less behavior data, such as new platform users or convenient users, the accuracy of the existing information push mode is low.

Disclosure of Invention

In view of this, embodiments of the present invention provide an information pushing method and apparatus, which can effectively improve information pushing accuracy for users with less behavior data, such as new platform users or convenient users.

In order to achieve the above object, according to an aspect of the embodiments of the present invention, there is provided an information pushing method, including:

screening a target object;

selecting anchor point information for the target object, and determining the category of the anchor point information;

acquiring behavior data related to the anchor point information, and calculating association indexes between the category of the anchor point information and various target categories included in the behavior data;

determining the category to be pushed according to the correlation index;

selecting information to be pushed from the categories to be pushed according to a preset pushing strategy;

and when a page processing request sent by the target object is received, pushing the information to be pushed for the target object.

Preferably, the screening target object includes:

screening one or more primary screening objects with a first type of interactive operation in a sampling period;

calculating the average interaction duration of the first type of interaction operation and the actual interaction duration for each primary screening object;

and screening a target object from one or more primary screened objects according to the average interaction duration and the actual interaction duration of the first type of interaction operation.

Preferably, calculating the average interaction duration of the first type of interaction operation includes:

counting a first usage time of a usage object of an application for the application and a first operation number of the usage object for the first type of interaction operation in the sampling period;

and calculating the average interaction duration for the first type of interaction operation according to the first use duration and the first operation times.

Preferably, calculating the actual interaction duration for each of the prescreened objects includes:

counting a second usage time of each of the preliminary screening objects for the application and a second operation number of each of the preliminary screening objects for the first type of interaction operation in the sampling period;

for each of the prescreened objects, performing: and calculating the actual interaction duration of the first type of interaction operation according to the second use duration and the second operation times of the primary screening object.

Preferably, screening the target object from one or more of the primary screening objects comprises:

performing, for each of the prescreened objects:

and judging whether the ratio of the actual interaction time length of the primary screening object to the average interaction time length meets a preset screening condition, and if so, determining the primary screening object as a target object.

Preferably, selecting anchor point information for the target object includes:

and determining target information which is executed by the target object and is aimed at the second type of interactive operation with the shortest distance to the current time, and taking the target information with the shortest distance to the current time as the anchor point information.

Preferably, the information pushing method further includes: setting a category tag for a plurality of information included in the application;

determining the category to which the anchor point information belongs, including:

and acquiring the category label included by the anchor point information.

Preferably, the acquiring of behavior data related to the anchor point information includes:

determining a complete time unit with the shortest distance to the current time;

and acquiring order data associated with the anchor point information in the complete time unit, wherein the order data is generated by a first type of interactive operation.

Preferably, the first and second electrodes are formed of a metal,

the order data includes: the execution object of the first type of interactive operation, a plurality of types of reference information belonging to the same order with the anchor point information, all orders to which the reference information belongs and the target class to which each type of the reference information belongs;

calculating an association index between the category to which the anchor point information belongs and a plurality of target categories included in the behavior data, including:

for each of the object classes, performing:

calculating the object number ratio of the target category and the contribution value of the target category in the complete time unit according to the execution object and all orders to which the reference information corresponding to the target category belongs;

and calculating the association index of the anchor point information belonging category and the target category in the complete time unit according to the object quantity ratio of the target category and the contribution value of the target category in the complete time unit.

Preferably, calculating the ratio of the number of objects in the target category includes:

counting the number of execution objects for executing the first type of interactive operation aiming at the reference information of the target category and the anchor point information simultaneously and the number of execution objects for executing the first type of interactive operation aiming at the reference information of the target category according to all orders to which the execution objects and the reference information corresponding to the target category belong;

and calculating the ratio of the number of the objects according to the statistical result.

Preferably, the information pushing method further includes:

for each of the object classes, performing:

counting correlation indexes corresponding to the complete time units;

and calculating a new association index for the target category according to the statistical result.

Preferably, calculating a new association index for the target category includes:

judging whether the total number of the associated indexes included in the statistical result is not less than the number of the preset indexes,

if so, selecting the maximum value of a preset number from the plurality of associated indexes; calculating a new correlation index by using the maximum value of the selected preset number;

otherwise, calculating a new correlation index by using the plurality of correlation indexes.

Preferably, the determining the category to be pushed includes:

performing descending arrangement on the correlation indexes corresponding to the multiple target categories;

selecting a plurality of associated indexes meeting the preset floating domain condition from the descending order arrangement result;

randomly selecting the correlation indexes of the pushing number of the preset floating domain from the correlation indexes meeting the preset floating domain condition;

and determining the target category corresponding to the randomly selected association index as the category to be pushed.

Preferably, the information pushing method further includes:

judging whether the number of the associated indexes meeting the preset floating domain condition is smaller than the preset floating domain pushing number, if so, selecting the associated indexes of the preset floating domain pushing number from the descending order result according to the descending order; otherwise, executing the step of randomly selecting the correlation index of the preset floating domain pushing number.

Preferably, the determining the category to be pushed further comprises:

calculating a median value according to the maximum value and the minimum value in the descending order arrangement result;

selecting the correlation index with the minimum difference value with the median from the descending order arrangement result;

and determining the target category corresponding to the correlation index with the minimum median difference value as the category to be pushed.

Preferably, the information pushing method further includes:

when the number of the associated indexes having the smallest difference value with the median is greater than 1,

and selecting a target category corresponding to the related index ranked in the front as the category to be pushed from the multiple related indexes with the minimum median difference value.

Preferably, the determining the category to be pushed further comprises:

selecting the last correlation index from the descending order arrangement result;

and determining the target category corresponding to the last correlation index as the category to be pushed.

Preferably, the information pushing method further includes:

setting a counter for the page where the information to be pushed is located;

after the information to be pushed is pushed for the target object, the counter executes counting and 1 adding operation;

and judging whether the current counting result of the counter is not less than a preset iteration threshold value, if so, restoring the current counting of the counter to an initial value, and stopping current pushing.

Preferably, after it is determined that the current counting result of the counter is not less than the preset iteration threshold, the method further includes:

and judging whether the target object meets a preset screening condition, if so, executing the step of selecting anchor point information for the target object.

Preferably, the information pushing method further includes:

and if the target object executes a second type of interactive operation aiming at the information to be pushed, stopping current pushing and executing the step of judging whether the target object meets the preset screening condition.

Preferably, the information pushing method further includes:

if the target object executes a third type of interactive operation aiming at the information to be pushed, reselecting the information to be pushed from the type to be pushed;

and pushing the information to be pushed after reselection for the target object.

Preferably, the information pushing method further includes:

and if the target object executes a third type of interaction operation aiming at the information to be pushed, the counter executes a count minus 1 operation.

In a second aspect, an embodiment of the present invention provides an information pushing apparatus, including: a screening unit, an index calculating unit and a pushing unit, wherein,

the screening unit is used for screening the target object;

the index calculation unit is used for selecting anchor point information for the target object and determining the category of the anchor point information; acquiring behavior data related to the anchor point information, and calculating a correlation index between the category to which the anchor point information belongs and a plurality of target categories included in the behavior data;

the pushing unit is used for determining the category to be pushed according to the category to which the anchor point information belongs and the associated indexes corresponding to the various target categories respectively; selecting information to be pushed from the categories to be pushed according to a preset pushing strategy; and when a page processing request sent by the target object is received, pushing the information to be pushed for the target object.

One embodiment of the above invention has the following advantages or benefits: according to the method, anchor point information is selected for a screened target object (the screened target object is a user with less behavior data such as a new platform user or a convenient user), the anchor point information is information which is concerned by the target object, the category to be pushed with larger relevance from the category to which the anchor point information belongs can be determined by calculating the relevance index between the category to which the anchor point information belongs and various target categories included by the behavior data, the determined category to be pushed is enabled to be matched with the requirement of the target object in a comparison mode, and the information to be pushed can be selected from the category to be pushed subsequently according to a preset pushing strategy to be pushed to the target object. The process of selecting the information to be pushed is based on the anchor point information, the anchor point information is information with strong interaction with the target object, and the requirements of users with less behavior data, such as platform new users or convenient users, can be reflected more truly, so that the accuracy of information pushing can be effectively improved for the users with less behavior data, such as platform new users or convenient users.

Further effects of the above-mentioned non-conventional alternatives will be described below in connection with specific 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 diagram of a main flow of an information push method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a main flow of screening a target object according to an embodiment of the present invention;

FIG. 3 is a diagram illustrating a main process of calculating a correlation index according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a main flow of calculating a new correlation indicator for a target class according to an embodiment of the present invention;

FIG. 5 is an illustration of a main flow of determining a category to be pushed according to one embodiment of the invention;

FIG. 6 is a schematic diagram of a main flow of determining a category to be pushed according to another embodiment of the present invention;

FIG. 7 is a schematic diagram of a main flow of determining a category to be pushed according to yet another embodiment of the present invention;

FIG. 8 is a diagram of anchor point information, mid-bit information, and excitation information, according to one embodiment of the invention;

fig. 9 is a schematic diagram of a main flow of an information pushing method according to another embodiment of the present invention;

fig. 10 is a schematic diagram of a main flow of an information push method according to still another embodiment of the present invention;

fig. 11 is a schematic diagram of the main units of an information pushing apparatus according to an embodiment of the present invention;

FIG. 12 is an exemplary system architecture diagram in which embodiments of the present invention may be employed;

fig. 13 is a schematic structural diagram of a computer system suitable for use in implementing a terminal device or server of an embodiment of the present 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 exemplary only. 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.

In marketing, users can be divided into four categories, economic users, moral users, personalized users, and convenience users. Wherein, economical user: the price sensitivity is high, the preference activity commodity and the discount commodity can be solid commodities such as clothes, shoes, hats, furniture and the like, and can also be virtual commodities such as rechargeable cards, video advertisements and the like; moral-type user: the brand goods are more inclined, and the public praise and the user evaluation are more concerned; the personalized user: the method has detailed standards for the attributes of the required commodities, and the personalized user often completes commodity selection by searching the attributes and the characteristics of the commodities in an e-commerce scene; in a video watching/video advertisement scene, the selection is often performed by searching for the theme, attribute, and the like of the video; the convenient user: it is hoped that shopping is easy or video recommendation/push is easy to obtain, judgments are not favored, and a large variety of searches are favored to select products with high comprehensive ranking or buy the products or watch videos according to recommendation/push, and the behaviors are irregular. Currently, an effective pushing method is not available for a convenient user or a new user lacking behavior data. The invention combines experience data to screen convenient users or new users lacking behavior data from various objects (such as commodity purchasing users, video watching users and the like). Because the relevance requirement of products purchased by or videos watched by the convenient users or new users lacking behavior data is not high, and the efficiency and the applicability are emphasized, based on the characteristics of the convenient users or the new users lacking behavior data, the invention designs an information pushing method for the users. Fig. 1 is an information pushing method according to an embodiment of the present invention, and as shown in fig. 1, the information pushing method may include the following steps:

step S101: screening a target object;

the target object may refer to a user name or a page login account corresponding to the convenient user or the new user lacking the behavior data, or may refer to a client or an application used by the convenient user or the new user lacking the behavior data.

Step S102: selecting anchor point information for a target object, and determining the category of the anchor point information;

the category is a category configured or allocated for information in advance, and for example, for a physical commodity, the category may include: meat, fruits, vegetables, coats, trousers, skirts, shoes, wardrobes, beds, ornaments and the like; in addition, for the physical goods, the categories may be: beef, mutton, pork, strawberry, blueberry, watermelon, banana, tomato, celery, Chinese cabbage, jeans, sport pants and the like can express the types of specific forms of commodities. For virtual goods, the categories may include: game rechargeable cards, communication rechargeable cards, etc.; for video information, its categories may include: fighting, living, sports, entertainment, live shopping, live game, secondary element, tertiary element and the like. For example, if the anchor point information is live shopping, the category to which the anchor point information belongs is live shopping, and for example, if the anchor point information is strawberry purchased by a user, the category to which the anchor point information belongs is fruit or strawberry; the anchor point information is a movie of a certain fighting subject with the watching time length exceeding the preset time length, and the anchor point information is a fighting category.

Step S103: acquiring behavior data related to anchor point information, and calculating association indexes between the category of the anchor point information and various target categories included in the behavior data;

the behavior data related to the anchor point information refers to the fact that the anchor point information has direct or indirect relation (for example, information related to other information belonging to the same order, the same order as the anchor point information, the same user browsed or watched at the same time as the anchor point information, or the same user browsed or watched at the same time as the other information). For example, the anchor point information is a product a purchased by the user, the other information simultaneously ordered with the product a is a product B and a product C, and the orders including any one of the product a, the product B, the product C, the product D, and the product E are behavior data related to the anchor point information product a. For another example, if the video advertisement 1 is anchor information, the user viewing/browsing the video advertisement 1 browses the video advertisement 2 at the same time, and the user viewing/browsing the video advertisement 2 browses the video advertisement 3 at the same time, the user information of viewing/browsing the video advertisement 1, the video advertisement 2, or the video advertisement 3, the browsing information of the user, and the like are behavior data related to the anchor information video advertisement 1.

Step S104: determining the category to be pushed according to the correlation index;

step S105: selecting information to be pushed from the categories to be pushed according to a preset pushing strategy;

the push policy may be: pushing is performed according to the sales volume, the viewing volume, the attention volume, the evaluation and the like from the categories to be pushed, for example, if the categories to be pushed are watermelons, watermelons with the highest sales volume or the best evaluation are selected as information to be pushed from all merchants selling the watermelons, and if the categories to be pushed are short-lived videos, short-lived videos with the highest viewing volume/the best evaluation are selected as information to be pushed.

Step S106: and pushing information to be pushed for the target object when a page processing request sent by the target object is received.

The page processing request can be a request for page refreshing, logging in/entering a page, and the like, and the information to be pushed for the target object can be pushed in the pushing position of the page according to the user name, the user login account number, the used client, and the like of the target object.

In an embodiment of the present invention, as shown in fig. 2, the specific implementation of the step S101 may include the following steps:

step S201: screening one or more primary screening objects with a first type of interactive operation in a sampling period;

the sampling period may be set accordingly according to the requirement, such as 7 days, 14 days, etc. The first type of interactive operation may refer to a purchasing operation, a viewing operation/browsing operation, and the duration of the viewing operation/browsing operation reaches a preset duration threshold, etc. For example, within 7 days of screening, there are users who have bought operations, and for example, within 7 days of screening, there are users who have watched video for a time period exceeding a preset time period threshold value.

As the screened target object is a convenient user or a new user lacking behavior data, tests show that the sampling period is generally not less than 7 days, and the target object can be screened more accurately.

Step S202: calculating the average interaction duration of the first type of interaction operation and the actual interaction duration for each prescreened object;

the specific implementation manner of calculating the average interaction duration of the first type of interaction operation may include: counting a first use duration of a use object of the application for the application and a first operation number of the use object for a first type of interactive operation in a sampling period; and calculating the average interactive duration for the first type of interactive operation according to the first use duration and the first operation times.

The application use object refers to a user who operates a client, a web page version application, and the like, such as a commodity purchasing user, a video viewing user, and the like. The first usage duration refers to the total usage duration of the application usage object to the client and the web page version application in the statistical period, the first operation frequency refers to the total frequency of the application usage object to perform the first type of interactive operation on the client and the web page version application in the statistical period, for example, the first type of interactive operation is a commodity purchasing operation, and the first operation frequency is the total frequency of the application usage object to perform the commodity purchasing operation on the client and the web page version application in the statistical period; for another example, if the first type of interaction operation is that the video watching time length is not lower than the preset time length threshold, the first operation frequency is the total frequency that the user object applied in the statistical period executes the video watching operation on the client and the web page version application, and the watching time length is not lower than the preset time length threshold.

The above-mentioned calculation of the average interaction duration for the first type of interaction operation can be obtained by the following calculation formula (1).

Wherein, T_zRepresenting the calculated average interactive duration of the first-class interactive operation; t is_yCharacterizing a first usage duration of a counted usage object of the application for the application; v. of_yA first number of operations of the usage object for the first type of interaction is characterized.

In this step, the specific implementation manner of calculating the actual interaction duration for each prescreened object may include: in a sampling period, counting a second use duration of each primary screening object for the application and a second operation frequency of each primary screening object for the first type of interactive operation; for each prescreening object, performing: and calculating the actual interactive duration of the first type of interactive operation according to the second use duration and the second operation times of the primary screening object.

The second use duration for the application is the total duration of the primary screening object using the client or the application webpage version in the sampling period. The second operation frequency of the prescreened object for the first type of interactive operation refers to the total frequency of the prescreened object for the first type of interactive operation in the sampling period, for example, the first type of interactive operation is a commodity purchasing operation, and the second operation frequency refers to the total frequency of purchasing operations of the prescreened object in the sampling period; the first type of interactive operation is video watching operation, the watching time length is not lower than a preset time length threshold value, and the second operation times refer to the total times that the watching time length of the video watching operation of the primary screening object in the sampling period is not lower than the preset time length threshold value in the sampling period;

specifically, the actual interaction duration for the first-type interaction operation can be calculated by the following calculation formula (2).

Wherein, T_piRepresenting the actual interaction time length of the first-class interaction operation of the calculated primary screening object i; t is_iRepresenting the second service time of the counted primary screening object i for the application; v. of_tiAnd characterizing the second operation times of the counted primary screening object i aiming at the first type of interactive operation.

Step S203: and screening the target object from one or more primary screening objects according to the average interaction duration and the actual interaction duration of the first type of interaction operation.

The specific implementation of step S203 may include: performing, for each of the prescreened objects:

judging whether the ratio of the actual interaction duration to the average interaction duration of the primary screening object meets a preset screening condition or not, and if so, determining the primary screening object as a target object; otherwise, determining that the primary screening object is not the target object.

The preset screening condition may be that a ratio of the actual interaction duration to the average interaction duration of the preliminarily screened object is not less than a preset ratio threshold. The ratio threshold value can be set or adjusted according to actual conditions. Such as a ratio threshold of 0.8.

For exampleThe ratio threshold for a certain e-commerce platform is 0.8. The first type of interaction is a purchase operation. 1000 registered users exist on the e-commerce platform, and 600 search results are obtained by searching for users who have purchase operation within 7 days in the history record in the log, namely the 600 results are primary screening objects. Calculating the average time length T spent on purchasing operation of the user according to the background data_zIf the cumulative service life of the client and the application web page version of the e-commerce platform within 7 days is 3000 hours and the cumulative purchase frequency is 9000 times, then T is determined_y＝3000，v_y＝9000，

Calculating the time length T spent on single purchasing behavior of each primary screening object in 600 primary screening objects_piTaking the user A as an example, the user ID of the user A is 0001, and the service life of the E-commerce platform in the 7 days is 2 hours T_i2, the number of purchases is 8 v_tiWhen it is equal to 8, then

Further calculating the ratio of the actual interaction duration to the average interaction duration of the primary screening object:

if the user A (user ID0001) meets the preset screening condition, the user A is determined to be the target object 0001. For another example, if the usage duration of the user b using the e-commerce platform in the 7 days is 3 hours and the purchase frequency is 1 in the 600 primary screening objects, the user b does not satisfy the preset screening condition, that is, the user b is not the target object and needs to be removed from the set of the primary screening objects.

In an embodiment of the present invention, the specific implementation manner of selecting anchor point information for a target object may include: and determining target information which is executed by the target object and is the shortest from the current time and is aimed at by the second type of interactive operation, and taking the target information as anchor point information. For example, for an e-commerce platform, the second type of interaction may be an interaction configured for the platform such as a shopping cart joining operation, a purchase operation, and the like. For the video playing platform, the second interactive operation may be a video watching operation, the video watching time length of which exceeds a preset time length threshold, a video clicking operation, the video clicking times of which reach a preset time threshold, and the like. The second type of interaction can be set accordingly according to actual requirements.

In an embodiment of the present invention, the information pushing method may further include: setting category labels for various information included in the application; the specific implementation manner for determining the category to which the anchor point information belongs may include: and acquiring the category label included by the anchor point information.

In the embodiment of the present invention, the specific implementation manner of acquiring the behavior data related to the anchor point information may include: determining a complete time unit with the shortest distance to the current time; and acquiring order data which is associated with the anchor point information in the complete time unit, wherein the order data is generated by the first type of interactive operation. The current time refers to the time recorded/displayed by the equipment when the equipment executes the information pushing method. A complete time unit may be set for the user according to actual requirements, such as the previous 24 hours corresponding to the current time (where the 24 hours may be set accordingly according to requirements). For example, if 24 hours are from 0 a morning to 0 a next morning and the current time is 8 a morning, the shortest complete time unit from the current time is from 0 a day before to 0 a day. For another example, if the current time is 8 am from 9 am to 9 am on the next day of a 24-hour day, the one complete time unit with the shortest distance to the current time is 9 am to 9 am on the previous day two days before the current time.

For the E-commerce platform, the order data is generated by purchase operation (first-type interaction operation), and the order data associated with the anchor point information is order data including order data of the order information and order data to which other information existing in the same order data with the anchor point information belongs, for example, the anchor point information is a product a purchased by a user, the other information simultaneously issued with the product a is a product B and a product C, and the order of any one of the product a, the product B, the product C, the product D and the product E belongs to the same order as the product B or the product C is order data related to the anchor point information product a.

For another example, for a video playing platform or a live broadcast platform, the order data is generated by a watching operation (a first kind of interactive operation), and the order data associated with the anchor information means that the video advertisement 1 is the anchor information, a user watching/browsing the video advertisement 1 browses the video advertisement 2 at the same time, a user watching/browsing the video advertisement 2 browses the video advertisement 3 at the same time, and the user information watching/browsing the video advertisement 1, the video advertisement 2, or the video advertisement 3, the browsing information of the user, and the like are anchor behavior data associated with the video advertisement 1.

In an embodiment of the present invention, the order data may include: the method comprises the steps that a first-class interactive operation execution object, multiple kinds of reference information belonging to the same order with anchor point information, all orders to which the reference information belongs and a target class to which each kind of reference information belongs are obtained; accordingly, as shown in fig. 3, a specific embodiment of calculating an association indicator between a category to which anchor point information belongs and a plurality of target categories included in behavior data may include: for each object class, the following steps S301 and S302 are performed:

step S301: calculating the object quantity ratio of the target category and the contribution value of the target category in a complete time unit according to all orders to which the reference information corresponding to the execution object and the target category belongs;

in this step, the specific implementation of calculating the ratio of the number of objects in the target category may include: counting the number of execution objects which execute the first type of interactive operation aiming at the reference information of the target category and the anchor point information at the same time and the number of execution objects which execute the first type of interactive operation aiming at the reference information of the target category according to all orders to which the execution objects and the reference information corresponding to the target category belong; and calculating the ratio of the number of the objects according to the statistical result.

The execution object refers to an execution object that performs a first type of interactive operation on the anchor point information and other information related to the anchor point information, for example, if the anchor point information is a product a and the information related to the anchor point information is a product B, the execution object refers to a user that performs a first type of interactive operation (for example, a purchase operation) on the product a or the product B; the target category is a category to which the product B belongs, and the product B is a reference object. For another example, if the anchor point information is video 1, and video 2 is associated with video 1, the execution object refers to a user or the like that performs a first type of interaction operation (for example, watching a video for a time period not less than a preset time period threshold value) on video 1 or video 2, the target category is a category to which video 2 belongs, and video 2 is a reference object.

Specifically, the object number proportion of the calculation target class can be calculated according to the following calculation formula (3).

Wherein r is_sThe number of objects representing the target class is proportional; n is a radical of_ijRepresenting the counted number of execution objects for executing the first-class interaction operation aiming at the reference information j and the anchor point information i of the target class; n is a radical of_jThe number of execution objects that characterize the execution of the first type of interaction with respect to the reference information j of the target class. For example, the anchor information is a commodity a, the reference information of the target category is a commodity B, and in a complete time unit shortest from the current time, 100 people perform a purchase operation on the commodity B (that is, the number of execution objects (purchase users) performing the first type of interaction operation (purchase operation) on the reference information of the target category is 100), and 30 people simultaneously purchase the commodity a and the commodity B (the number of execution objects simultaneously performing the first type of interaction operation on the reference information of the target category and the anchor information is 30), the number of objects of the target category is defined as the ratio of the number of objects of the target category to the number of objects of the target category

In addition, the contribution value of the target category in the complete time unit can be calculated according to the number of reference information in the order which simultaneously includes the anchor point information and the reference information of the target category in the complete time unit and the total number of reference information included in all order data in the complete time unit.

Specifically, the contribution value of the target class in the complete time unit may be calculated according to the following calculation formula (4).

Wherein r is_xjRepresenting a contribution value of a target class x to which the reference information j belongs; f_ijRepresenting the contribution amount of reference information j in an order which simultaneously comprises anchor point information i and reference information j of a target category in a complete time unit; f_jThe total contribution of the reference information j included in all order data within a complete time unit is characterized. For example, for a shopping scenario of the e-commerce platform, the reference information is a product B, the sales amount generated in a complete time unit is 10000 (in the complete time unit, the total contribution amount of the reference information j included in all order data), and a user who purchases the anchor information product a and the product B at the same time contributes 1000 of the sales amount (in the complete time unit, the order including the anchor information i and the reference information j of the target category includes the contribution amount of the reference information j), the contribution value of the target category x to which the reference information j belongs is 10000

For another example, for a video playing platform, the reference information is video 2, the video playing amount generated in a complete time unit is 0000 (in the complete time unit, the total contribution amount of reference information j included in all order data), and the user playing anchor information video 1 and video 2 at the same time contributes 1000 of the video playing amount (in the complete time unit, the order including anchor information i and reference information j of the target category includes the contribution amount of reference information j), the contribution value of the target category x to which the reference information j belongs is 0000, and the contribution value of the target category x to which the reference information j belongs is 0000 (in the complete time unit, the contribution amount of reference information j is included in the order including anchor information i and reference information j of the target category at the same time)

Step S302: and calculating the association index of the anchor point information belonging category and the target category in the complete time unit according to the object quantity ratio of the target category and the contribution value of the target category in the complete time unit.

Specifically, the association index of the category to which the anchor point information belongs and the target category in the complete time unit can be calculated according to the following calculation formula (5).

Wherein d is_ijRepresenting the association index of the category to which the anchor point information i belongs and the target category to which the reference information j belongs in the complete time unit; r is_sThe number of the objects of the target category to which the reference information j belongs, which is calculated by the representation calculation formula (3), is in proportion; r is_xjThe contribution value of the target category x to which the reference information j calculated by the formula (4) belongs is calculated.

Namely: the two-dimensional coordinate graph may be constructed based on the ratio of the number of objects of the target class to which the reference information j belongs calculated by the calculation formula (3) and the contribution value of the target class x to which the reference information j belongs calculated by the calculation formula (4), where r is_sAnd r_xjRespectively, an abscissa and an ordinate in the two-dimensional coordinate graph. Then the correlation index is r_sAnd r_xjThe distance between the formed coordinate point and the origin in the two-dimensional coordinate graph is larger, which indicates that the target category corresponding to the reference information is more strongly associated with the category corresponding to the anchor point information.

For example, r calculated for the previous embodiment_s＝0.3，r_xj＝0.1，

Then

It should be noted that, in order to accurately calculate the object of the category to which the anchor point information belongs and the object of the reference informationThe correlation index of the object class in the complete time unit, d_ijIt may be set to count once per complete time unit or once every 24 hours, i.e. a number of times d is counted in one sampling period_ij。

The high-value reference information can be prevented from being overlooked by calculating the target number ratio of the target category by the calculation formula (3) and the contribution value calculated by the calculation formula (4). For example, 76% of users who purchase toothbrushes (anchor information) buy toothpaste (reference information), and r thereof_s0.6, very high, but the average unit price of the toothpaste purchased by the users purchased at the same time is 1 yuan, and the average unit price of the toothpaste is 10 yuan for the rest 40 percent_xjAnd is not high. On the contrary, a small part of users who buy the toothbrush (anchor information) buy the high-end tooth cup (reference information), the average selling price of the high-end tooth cup is 500 yuan, and r corresponding to the high-end tooth cup_s0.1, but more users buy low end cups, e.g. r_s0.9, the product of the high-end tooth cup is r_sLow, but it cannot be said that it is of no value and is not strongly correlated, so r is required_xjThis indicator avoids this situation. To better find the target class for the target object.

In addition, d is obtained by the above-mentioned multiple calculation_ijCan be stored in the data set K, i.e.

K＝{d_ij1，d_ij2，...，d_ijnIn which d is_ijnCharacterizing d calculated at the nth time in a sampling period_ij. For example, if the aforementioned 0.316 is obtained by the second calculation, d is_ij20.316. In order to reduce the influence of holiday and seasonal factors on the product relevance, the relevance index between the category and the target category of the anchor point information is accurately obtained.

In an embodiment of the present invention, the information pushing method may further include: for each target class, performing: counting correlation indexes corresponding to a plurality of complete time units; and calculating a new correlation index for the target category according to the statistical result.

Specifically, as shown in fig. 4, the calculating of the new association index for the target category may include the following steps:

step S401: judging whether the total number of the associated indexes included in the statistical result is not less than the number of preset indexes, if so, executing the step S402; otherwise, go to step S403;

for example, if the number of preset indexes is t, the statistical result is d included in the data set K_ijIf the total number n is greater than t, executing step S402; if n ≦ t, step S403 is performed.

Step S402: selecting a maximum value of a preset number from a plurality of correlation indexes; calculating a new correlation index by using the maximum value of the selected preset number, and finishing the current process;

the specific implementation manner of the step can be as follows: and performing descending arrangement on the plurality of correlation indexes, taking t correlation indexes from high to low, and calculating an average value of the t correlation indexes. That is, it is calculated by the following calculation formula (6).

Wherein d is_ij' characterizing the new correlation index calculated by the calculation formula (6);

d_ij1，d_ij2，...d_ijtthe characterization takes t items of correlation indexes from high to low from a data set K; n represents the total number of the associated indexes included in the data set K.

Step S403: and calculating a new correlation index by using the plurality of correlation indexes.

The specific implementation manner of the step can be as follows: and taking all the associated indexes in the data set K, and calculating an average value of the n associated indexes. That is, it is calculated by the following calculation formula (7).

Wherein d is_ij"characterize the new correlation index calculated by calculation formula (7);

d_ij1，d_ij2，...d_ijnall correlation indexes included in the characterization data set K; n represents the total number of the associated indexes included in the data set K; and t represents the number of preset indexes.

It should be noted that the value of t is only related to the calculation accuracy, so as to increase the robustness of the algorithm. For example, if t is 3, the association index set of the product category a and the product category B is

{0.318, 0.316, 0.280, 0.350, 0.322, 0.316, 0}, the first three items in descending order

0.350, 0.322, 0.318, then

Similarly, calculating the associated indexes of the category to which the anchor point information belongs and the categories to which the rest of reference information belongs, and arranging the associated indexes into a set G, wherein G is { d ═ d_i1′/d_i1″，d_i2′/d_i2″，...，d_im′/d_im". Wherein d is_i1′/d_i1"characterize the association index between the category to which the anchor point information belongs and the target category to which the 1 st reference information belongs; d_im′/d_im"characterize the association index between the category to which the anchor point information belongs and the target category to which the mth reference information belongs.

In the embodiment of the present invention, as shown in fig. 5, determining the category to be pushed may include the following steps:

step S501: performing descending order arrangement on the correlation indexes corresponding to the various target categories;

step S502: selecting a plurality of correlation indexes meeting the preset floating domain condition from the descending order arrangement result;

a floating domain refers to a class of objects that float in the vicinity of anchor point information, i.e., d_ij′/d_ij"comparatively large object class. For example, regarding the sorted result in step S501, a plurality of related indexes in which the target category corresponding to the top 20% of the related indexes in the sorted result is the preset floating domain condition are defined, where the 20% is an adjustable value, and may also be 30%, 40%, and the like.

Step S503: randomly selecting the correlation indexes of the pushing number of the preset floating domain from the correlation indexes meeting the preset floating domain condition;

for example, if the number of the associated indexes meeting the preset floating domain condition is 10, and the number of the preset floating domain push is 3, then 3 associated indexes are randomly selected from the 10 associated indexes meeting the preset floating domain condition.

Step S504: and determining the target category corresponding to the randomly selected association index as the category to be pushed.

In an embodiment of the present invention, the information pushing method may further include: judging whether the number of the associated indexes meeting the preset floating domain condition is smaller than the preset floating domain pushing number, if so, selecting the associated indexes of the preset floating domain pushing number from the descending order result according to the descending order; otherwise, the step of randomly selecting the correlation indexes of the preset floating domain pushing number is executed. For example, if the number of the association indexes meeting the preset floating domain condition selected in step S502 is 2, and if the number of the association indexes is 3, the association indexes ranked in the top 3 are directly selected from the results sorted in step S501, and the target category corresponding to the selected association index is the category to be pushed.

In the embodiment of the present invention, as shown in fig. 6, determining the category to be pushed may further include the following steps:

step S601: calculating a median value according to the maximum value and the minimum value in the descending order arrangement result;

this step can be calculated according to the calculation formula (8).

Wherein d is_zThe median value calculated in the step S601 is represented; max (g) characterize a maximum value in a set comprising correlation indexes corresponding to a plurality of target classes; min (G) characterizing the minimum value in the set of correlation indexes corresponding to the target classes.

Step S602: selecting the correlation index with the minimum difference value with the median value from the descending order arrangement result;

μ＝|d_z-d_ij′|

wherein mu represents the difference between the correlation index and the median in the set G; d_zThe median value calculated in the characterization step S601; d_ijRepresenting an association index between a category to which anchor point information i belongs and a target category of reference information j; understandably, d is_ij' alternative to d_ij″。

Step S603: and determining the target category corresponding to the correlation index with the minimum median difference value as the category to be pushed.

In the embodiment of the present invention, the information pushing method may further include: and when the number of the associated indexes with the minimum median difference is larger than 1, selecting the target category corresponding to the associated index in the front sequence as the category to be pushed from the associated indexes with the minimum median difference.

In addition, in order to improve the access amount of the information with low access degree, in the embodiment of the present invention, as shown in fig. 7, determining the category to be pushed may further include the following steps:

step S701: selecting the last correlation index from the descending order arrangement result;

step S702: and determining the target category corresponding to the last correlation index as the category to be pushed.

For example, the sorted set G ═ 0.33, 0.32, 0.30, 0.30,. and 0}, assuming that the set has 20 association indexes, where the floating domain condition is that the association indexes sorted in the top 20% are taken, the floating domain of the anchor point product is 4, and the target class corresponding to the set G from left to right is 0.33 corresponding to the class B, 0.32 corresponding to the class C, 0.30 corresponding to the class D, and 0.30 corresponding to the class E, and three items are randomly extracted, and assuming that three items of the class C, D, E are extracted, it is determined that the target class to be pushed is C, D, E through the floating domain. And if the target class corresponding to the last correlation index 0 is F, using the F as an excitation node to be classified into the class to be pushed. Median value

And if two adjacent correlation indexes in the set G are respectively 0.170 and 0.160, and the difference between the correlation indexes and the median is 0.05, taking the target category corresponding to 0.170 as the category to be pushed, and if more than one item is 0.170, randomly selecting from the 0.170. And assuming that the target category selected by the median is Q, the iteration sequence of the categories to be pushed is C-D-E-F-Q in sequence.

Specifically, anchor point information, midpoint information, and excitation information as shown in fig. 8. Wherein, the anchor point information corresponds to a floating domain, and the floating domain and the category to be pushed selected from the floating domain are obtained by the embodiment shown in fig. 5; the midpoint information is obtained by the embodiment shown in fig. 6 described above; the excitation information is obtained from the embodiment shown in fig. 7. Namely, the information category to be pushed is composed of a target category in the floating domain, a target category corresponding to the middle point information and a target category corresponding to the excitation information, so that the access amount of the target category corresponding to the middle point information and the target category corresponding to the excitation information is increased.

In the embodiment of the present invention, as shown in fig. 9, the information pushing method may further include the following steps:

step S901: setting a counter for a page where information to be pushed is located;

step S902: after pushing information to be pushed for a target object, a counter executes counting and 1 adding operation;

step S903: judging whether the current counting result of the counter is not less than a preset iteration threshold, if so, executing a step S904; otherwise, executing step S902;

step S904: and restoring the current count of the counter to the initial value, and stopping the current pushing.

By the aid of the judging process, the situation that the object which is not the target object is pushed continuously in the sampling period according to the information pushing method provided by the embodiment of the invention can be avoided, resource waste is avoided, and meanwhile, the pushing accuracy and precision can be guaranteed.

In an embodiment of the present invention, after determining that the current counting result of the counter is not less than the preset iteration threshold, the method may further include: and judging whether the target object meets a preset screening condition, if so, executing the step of selecting anchor point information for the target object.

In the embodiment of the present invention, the information pushing method may further include: and if the target object executes the second type of interactive operation aiming at the information to be pushed, stopping the current pushing and executing the step of judging whether the target object meets the preset screening condition. For the e-commerce platform, the second type of interaction operation refers to a commodity plus shopping operation, a commodity purchasing operation and the like.

In the embodiment of the present invention, the information pushing method may further include: if the target object executes a third type of interactive operation aiming at the information to be pushed, reselecting the information to be pushed from the type to be pushed; and pushing the reselected information to be pushed for the target object. The counter performs a count minus 1 operation. The third type of interactive operation is a search operation, a page browsing operation, a praise operation, a collection operation and the like.

In order to clearly illustrate the role of the counter in the information pushing process, a specific embodiment is further described below. As shown in fig. 10, this embodiment may include the steps of:

step S1001: setting a counter for a page where information to be pushed is located;

step S1002: after pushing information to be pushed for a target object, a counter executes counting and 1 adding operation;

step S1003: judging whether interactive operation exists, if so, executing step 1007; when the interactive operation is the third type of interactive operation, executing step S1004; otherwise, go to step S1005;

step S1004: the counter executes the operation of counting and subtracting 1, and reselects the information to be pushed from the category to be pushed; pushing the newly selected information to be pushed for the target object, pushing the newly selected information to be pushed, and executing the step S1002;

step S1005: judging whether the current counting result of the counter is greater than or equal to a preset iteration threshold value or not, and if so, executing a step S1006; otherwise, go to step S1010;

step S1006: restoring the current count of the counter to the initial value;

step S1007: stopping current pushing;

step S1008: judging whether the target object meets a preset screening condition, if so, executing a step S1009; otherwise, directly stopping pushing;

step S1009: executing the step of selecting anchor point information for the target object and the subsequent step of selecting push information, and ending the current process;

it is understood that the step of selecting the push information subsequently is the steps from step S102 to step S106 in the embodiment 1 described above in fig. 1 and the operation of the embodiment related to these steps.

Step S1010: after the page is refreshed, the information to be pushed is pushed again, and step S1002 is executed.

As shown in fig. 11, an embodiment of the present invention provides an information pushing apparatus 1100, where the information pushing apparatus 1100 may include: a filtering unit 1101, an index calculation unit 1102, and a pushing unit 1103, wherein,

a screening unit 1101 for screening a target object;

the index calculation unit 1102 is configured to select anchor point information for the target object, and determine a category to which the anchor point information belongs; acquiring behavior data related to anchor point information, and calculating association indexes between the category of the anchor point information and various target categories included in the behavior data;

the pushing unit 1103 is configured to determine a category to be pushed according to the category to which the anchor point information belongs and the association indexes respectively corresponding to the multiple target categories; selecting information to be pushed from the categories to be pushed according to a preset pushing strategy; and pushing information to be pushed for the target object when a page processing request sent by the target object is received.

In this embodiment of the present invention, the screening unit 1101 is configured to screen one or more prescreened objects having a first type of interaction operation in one sampling period; calculating the average interaction duration of the first type of interaction operation and the actual interaction duration for each prescreened object; and screening the target object from one or more primary screening objects according to the average interaction duration and the actual interaction duration of the first type of interaction operation.

In this embodiment of the present invention, the screening unit 1101 is further configured to count, in a sampling period, a first usage duration of a usage object of the application for the application and a first operation number of the usage object for a first class of interaction operations; and calculating the average interactive duration for the first type of interactive operation according to the first use duration and the first operation times.

In this embodiment of the present invention, the screening unit 1101 is further configured to count a second usage duration of each prescreened object for the application and a second operation number of each prescreened object for the first type of interaction operations in a sampling period; for each prescreening object, performing: and calculating the actual interactive duration of the first type of interactive operation according to the second use duration and the second operation times of the primary screening object.

In this embodiment of the present invention, the screening unit 1101 is further configured to, for each prescreened object: and judging whether the ratio of the actual interaction time length to the average interaction time length of the primary screening object meets a preset screening condition, and if so, determining the primary screening object as a target object.

In this embodiment of the present invention, the index calculating unit 1102 is configured to determine target information, which is the shortest distance from the current time and is targeted by the target object to perform the second type of interactive operation, and use the target information, which is the shortest distance from the current time, as the anchor point information.

In this embodiment of the present invention, the index calculating unit 1102 is further configured to set category labels for a plurality of information included in the application; and acquiring the category label included by the anchor point information.

In this embodiment of the present invention, the index calculating unit 1102 is further configured to determine a complete time unit that is shortest from the current time; acquiring order data associated with anchor point information in a complete time unit, wherein the order data is generated by a first type of interaction operation.

In an embodiment of the present invention, the order data includes: the method comprises the steps that an execution object of a first type of interactive operation, multiple kinds of reference information belonging to the same order with anchor point information, all orders to which the reference information belongs and a target class to which each kind of reference information belongs are obtained; an index calculation unit 1102, further configured to perform, for each target category: calculating the object quantity ratio of the target category and the contribution value of the target category in a complete time unit according to all orders to which the reference information corresponding to the execution object and the target category belongs; and calculating the association index of the anchor point information belonging category and the target category in the complete time unit according to the object quantity ratio of the target category and the contribution value of the target category in the complete time unit.

In this embodiment of the present invention, the index calculation unit 1102 is further configured to count, according to all orders to which the reference information corresponding to the execution object and the target category belongs, the number of execution objects that perform the first type of interactive operation simultaneously with respect to the reference information of the target category and the anchor point information, and the number of execution objects that perform the first type of interactive operation with respect to the reference information of the target category; and calculating the ratio of the number of the objects according to the statistical result.

In this embodiment of the present invention, the index calculating unit 1102 is further configured to, for each target class, perform: counting correlation indexes corresponding to a plurality of complete time units; and calculating a new association index for the target category according to the statistical result.

In this embodiment of the present invention, the index calculating unit 1102 is further configured to determine whether a total number of associated indexes included in the statistical result is not less than a preset number of indexes, and if so, select a maximum value of the preset number from the multiple associated indexes; calculating a new correlation index by using the maximum value of the selected preset number; otherwise, calculating a new correlation index by using the plurality of correlation indexes.

In this embodiment of the present invention, the pushing unit 1103 is configured to sort association indexes corresponding to multiple target categories in a descending order; selecting a plurality of correlation indexes meeting preset floating domain conditions from the descending order arrangement result; randomly selecting the correlation indexes of the pushing number of the preset floating domain from the correlation indexes meeting the preset floating domain condition; and determining the target category corresponding to the randomly selected association index as the category to be pushed.

In this embodiment of the present invention, the pushing unit 1103 is further configured to determine whether the number of the associated indicators meeting the preset floating domain condition is smaller than the preset floating domain pushing number, and if so, select the associated indicators of the preset floating domain pushing number from the descending order result according to a descending order; otherwise, the step of randomly selecting the correlation indexes of the preset floating domain pushing number is executed.

In this embodiment of the present invention, the pushing unit 1103 is further configured to calculate a median value according to the maximum value and the minimum value in the descending order; selecting the correlation index with the minimum difference value with the median value from the descending order arrangement result; and determining the target category corresponding to the correlation index with the minimum median difference value as the category to be pushed.

In this embodiment of the present invention, the pushing unit 1103 is further configured to select, when the number of the associated indicators with the minimum median difference is greater than 1, a target category corresponding to the top-ranked associated indicator from the plurality of associated indicators with the minimum median difference as the category to be pushed.

In this embodiment of the present invention, the pushing unit 1103 is further configured to select a last correlation index from the results of the descending order arrangement; and determining the target class corresponding to the last correlation index as the class to be pushed.

In this embodiment of the present invention, as shown in fig. 11, the pushing unit 1103 is further configured to set a counter 1104 for a page where information to be pushed is located; after the information to be pushed is pushed for the target object, the counter 1104 performs a count plus 1 operation; and judging whether the current counting result of the counter 1104 is not less than a preset iteration threshold, if so, restoring the current counting of the counter 1104 to an initial value, and stopping current pushing.

In this embodiment of the present invention, the index calculating unit 1103 is further configured to determine whether the target object meets a preset filtering condition, and if yes, perform a step of selecting anchor point information for the target object.

In this embodiment of the present invention, the index calculation unit 1103 is further configured to, if the target object performs the second type of interaction operation on the information to be pushed, perform a step of determining whether the target object meets a preset screening condition.

In this embodiment of the present invention, the pushing unit 1103 is further configured to, if the target object performs a third type of interaction operation on the information to be pushed, reselect the information to be pushed from the category to be pushed; and pushing the reselected information to be pushed for the target object.

In this embodiment of the present invention, the pushing unit 1103 is further configured to control the counter 1104 to perform a count minus 1 operation if the target object performs a third type of interaction operation for the information to be pushed.

Fig. 12 shows an exemplary system architecture 1200 of an information pushing method or an information pushing apparatus to which an embodiment of the present invention can be applied.

As shown in fig. 12, the system architecture 1200 may include terminal devices 1201, 1202, 1203, a network 1204 and a server 1205. Network 1204 is the medium used to provide communication links between terminal devices 1201, 1202, 1203 and server 1205. Network 1204 may include various types of connections, such as wire, wireless communication links, or fiber optic cables, to name a few.

The user may use the terminal devices 1201, 1202, 1203 to interact with the server 1205 through the network 1204 to receive or send messages, etc. The terminal devices 1201, 1202, 1203 may have installed thereon various communication client applications, such as a shopping-like application, a web browser application, a search-like application, an instant messaging tool, a mailbox client, social platform software, etc. (by way of example only).

Terminal devices 1201, 1202, 12X03 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablets, laptop portable computers, desktop computers, and the like.

The server 1205 may be a server that provides various services, such as a background management server (for example only) that supports shopping-type websites browsed by users using the terminal devices 1201, 1202, 1203. The background management server may analyze and perform other processing on data and the like generated by the user through the terminal device through the interactive operation, and feed back a processing result (for example, information to be pushed — only an example) to the terminal device.

It should be noted that the information pushing method provided by the embodiment of the present invention is generally executed by the server 1205, and accordingly, the information pushing apparatus is generally disposed in the server 1205.

It should be understood that the number of terminal devices, networks, and servers in fig. 12 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

Referring now to FIG. 13, shown is a block diagram of a computer system 1300 suitable for use with a terminal device implementing an embodiment of the present invention. The terminal device or the server shown in fig. 13 is only an example, and should not bring any limitation to the functions and the range of use of the embodiment of the present invention.

As shown in fig. 13, the computer system 1300 includes a Central Processing Unit (CPU)1301 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)1302 or a program loaded from a storage section 1308 into a Random Access Memory (RAM) 1303. In the RAM 1303, various programs and data necessary for the operation of the system 1300 are also stored. The CPU 1301, the ROM 1302, and the RAM 1303 are connected to each other via a bus 1304. An input/output (I/O) interface 1305 is also connected to bus 1304.

The following components are connected to the I/O interface 1305: an input portion 1306 including a keyboard, a mouse, and the like; an output section 1307 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 1308 including a hard disk and the like; and a communication section 1309 including a network interface card such as a LAN card, a modem, or the like. The communication section 1309 performs communication processing via a network such as the internet. A drive 1310 is also connected to the I/O interface 1305 as needed. A removable medium 1311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1310 as necessary, so that a computer program read out therefrom is mounted into the storage portion 1308 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 embodiments, the computer program may be downloaded and installed from a network through communications component 1309 and/or installed from removable media 1311. The computer program executes the above-described functions defined in the system of the present invention when executed by a Central Processing Unit (CPU) 1301.

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 units described in the embodiments of the present invention may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes a filtering unit, an index calculation unit, and a pushing unit. Where the names of these units do not in some cases constitute a limitation on the units themselves, for example, a screening unit may also be described as a "unit of a screening target object".

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: screening a target object; selecting anchor point information for a target object, and determining the category of the anchor point information; acquiring behavior data related to anchor point information, and calculating association indexes between the category of the anchor point information and various target categories included in the behavior data; determining a category to be pushed according to the category to which the anchor point information belongs and the associated indexes respectively corresponding to the various target categories; selecting information to be pushed from categories to be pushed according to a preset pushing strategy; and pushing the information to be pushed for the target object when responding to the received page processing request sent by the target object.

According to the technical scheme of the embodiment of the invention, the anchor point information is selected aiming at the screened target object (the screened target object is a user with less behavior data such as a new platform user or a convenient user and the like), the anchor point information is the information which is concerned by the target object, the class to be pushed with larger relevance of the class to which the anchor point information belongs can be determined by calculating the relevance index between the class to which the anchor point information belongs and various target classes included in the behavior data, the determined class to be pushed can be matched with the requirement of the target object in a comparison manner, and the information to be pushed can be selected from the class to be pushed according to the preset pushing strategy to be pushed to the target object. The process of selecting the information to be pushed is based on the anchor point information, the anchor point information is information which is strongly interacted with the target object, and the requirements of users with less behavior data, such as platform new users or convenient users, can be reflected more truly, so that the accuracy of information pushing can be effectively improved for the users with less behavior data, such as platform new users or convenient users.

The above-described embodiments should not be construed as limiting the scope of the invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations, and substitutions may 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 (25)

1. An information pushing method, comprising:

screening a target object;

selecting anchor point information for the target object, and determining the category of the anchor point information;

acquiring behavior data related to the anchor point information, and calculating association indexes between the category to which the anchor point information belongs and various target categories included in the behavior data;

determining the category to be pushed according to the correlation index;

selecting information to be pushed from the categories to be pushed according to a preset pushing strategy;

and when a page processing request sent by the target object is received, pushing the information to be pushed for the target object.

2. The information pushing method according to claim 1, wherein the screening the target object includes:

screening one or more primary screening objects with a first type of interactive operation in a sampling period;

calculating the average interaction duration of the first type of interaction operation and the actual interaction duration for each prescreened object;

and screening a target object from one or more primary screened objects according to the average interaction duration and the actual interaction duration of the first type of interaction operation.

3. The information pushing method according to claim 2, wherein calculating the average interaction duration of the first type of interaction operation includes:

counting a first use duration of a use object of an application for the application and a first operation number of the use object for the first type of interaction operation in the sampling period;

and calculating the average interactive duration for the first type of interactive operation according to the first using duration and the first operation times.

4. The information pushing method according to claim 2, wherein calculating the actual interaction duration for each of the prescreened objects comprises:

counting a second usage duration of each of the prescreened objects for the application and a second operation number of each of the prescreened objects for the first type of interaction operation in the sampling period;

for each of the prescreened objects, performing: and calculating the actual interactive duration of the first type of interactive operation according to the second use duration and the second operation times of the primary screening object.

5. The information pushing method according to claim 2, wherein the screening of the target object from the one or more primary screened objects comprises:

performing, for each of the prescreened objects:

and judging whether the ratio of the actual interaction time length to the average interaction time length of the primary screening object meets a preset screening condition, if so, determining the primary screening object as a target object.

6. The information push method according to claim 1, wherein selecting anchor point information for the target object comprises:

and determining target information which is executed by the target object and is the shortest from the current time and is aimed at the second type of interactive operation, and taking the target information as the anchor point information.

7. The information pushing method according to claim 1,

further comprising: setting a category label for a plurality of information included in the application;

determining the category to which the anchor point information belongs, including:

and acquiring the category label included by the anchor point information.

8. The information pushing method according to claim 1, wherein acquiring behavior data related to the anchor point information includes:

determining a complete time unit with the shortest distance to the current time;

and acquiring order data associated with the anchor point information in the complete time unit, wherein the order data is generated by a first type of interactive operation.

9. The information pushing method according to claim 8,

the order data includes: the execution object of the first type of interactive operation, a plurality of types of reference information belonging to the same order with the anchor point information, all orders to which the reference information belongs, and the target category to which each type of the reference information belongs;

calculating an association index between the category to which the anchor point information belongs and a plurality of target categories included in the behavior data, including:

for each of the object classes, performing:

calculating the object quantity ratio of the target category and the contribution value of the target category in the complete time unit according to all orders to which the reference information corresponding to the execution object and the target category belongs;

and calculating the association index of the anchor point information belonging category and the target category in the complete time unit according to the object quantity ratio of the target category and the contribution value of the target category in the complete time unit.

10. The information pushing method according to claim 9, wherein calculating the ratio of the number of objects in the target category comprises:

counting the number of execution objects for executing the first type of interactive operation aiming at the reference information of the target category and the anchor point information simultaneously and the number of execution objects for executing the first type of interactive operation aiming at the reference information of the target category according to all orders to which the execution objects and the reference information corresponding to the target category belong;

and calculating the ratio of the number of the objects according to the statistical result.

11. The information pushing method according to claim 9, further comprising:

for each of the object classes, performing:

counting correlation indexes corresponding to the complete time units;

and calculating a new association index for the target category according to the statistical result.

12. The information push method according to claim 11, wherein calculating a new association index for the target category comprises:

judging whether the total number of the associated indexes included in the statistical result is not less than the preset index number,

if so, selecting the maximum value of a preset number from the plurality of associated indexes; calculating a new correlation index by using the maximum value of the selected preset number;

otherwise, calculating a new correlation index by using the plurality of correlation indexes.

13. The information pushing method according to claim 1 or 11, wherein determining the category to be pushed comprises:

performing descending arrangement on the correlation indexes corresponding to the multiple target categories;

selecting a plurality of correlation indexes meeting the preset floating domain condition from the descending order arrangement result;

randomly selecting the correlation indexes of the pushing number of the preset floating domain from the correlation indexes meeting the preset floating domain condition;

and determining the target category corresponding to the randomly selected association index as the category to be pushed.

14. The information pushing method according to claim 13, further comprising:

judging whether the number of the associated indexes meeting the preset floating domain condition is smaller than the preset floating domain pushing number, if so, selecting the associated indexes of the preset floating domain pushing number from the descending order result according to the descending order; otherwise, the step of randomly selecting the correlation indexes of the preset floating domain pushing number is executed.

15. The information pushing method according to claim 13, wherein determining the category to be pushed further comprises:

calculating a median value according to the maximum value and the minimum value in the descending order arrangement result;

selecting the correlation index with the minimum difference value with the median from the descending order arrangement result;

and determining the target category corresponding to the correlation index with the minimum median difference value as the category to be pushed.

16. The information pushing method according to claim 15, further comprising:

when the number of the associated indexes having the smallest difference value with the median is greater than 1,

and selecting a target category corresponding to the prior relevance index as the category to be pushed from the relevance indexes with the minimum median difference value.

17. The information pushing method according to claim 13, wherein determining the category to be pushed further comprises:

selecting the last correlation index from the descending order arrangement result;

and determining the target category corresponding to the last correlation index as the category to be pushed.

18. The information pushing method according to claim 1, further comprising:

setting a counter for the page where the information to be pushed is located;

after the information to be pushed is pushed for the target object, the counter executes counting and 1 adding operation;

and judging whether the current counting result of the counter is not less than a preset iteration threshold value, if so, restoring the current counting of the counter to an initial value, and stopping current pushing.

19. The information pushing method according to claim 18, further comprising, after determining that the current counting result of the counter is not less than a preset iteration threshold:

and judging whether the target object meets a preset screening condition, if so, executing the step of selecting anchor point information for the target object.

20. The information pushing method according to claim 19, further comprising:

and if the target object executes a second type of interactive operation aiming at the information to be pushed, stopping current pushing and executing the step of judging whether the target object meets the preset screening condition.

21. The information pushing method according to claim 19, further comprising:

if the target object performs a third type of interaction operation for the information to be pushed,

reselecting information to be pushed from the categories to be pushed;

and pushing the information to be pushed after reselection for the target object.

22. The information pushing method according to claim 21, further comprising:

and if the target object executes a third type of interactive operation aiming at the information to be pushed, the counter executes a count minus 1 operation.

23. An information pushing apparatus, comprising: a screening unit, an index calculation unit and a pushing unit, wherein,

the screening unit is used for screening the target object;

the index calculation unit is used for selecting anchor point information for the target object and determining the category of the anchor point information; acquiring behavior data related to the anchor point information, and calculating association indexes between the category to which the anchor point information belongs and various target categories included in the behavior data;

the pushing unit is used for determining a category to be pushed according to the category to which the anchor point information belongs and the associated indexes corresponding to the various target categories respectively; selecting information to be pushed from the categories to be pushed according to a preset pushing strategy; and when a page processing request sent by the target object is received, pushing the information to be pushed for the target object.

24. An information-pushing 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 1-22.

25. 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 1-22.

Images