CN110753247A - Information push strategy generation system and method - Google Patents
Information push strategy generation system and method Download PDFInfo
- Publication number
- CN110753247A CN110753247A CN201910994371.9A CN201910994371A CN110753247A CN 110753247 A CN110753247 A CN 110753247A CN 201910994371 A CN201910994371 A CN 201910994371A CN 110753247 A CN110753247 A CN 110753247A
- Authority
- CN
- China
- Prior art keywords
- user
- behavior data
- content
- preference
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
- H04N21/252—Processing of multiple end-users' preferences to derive collaborative data
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
- H04N21/25866—Management of end-user data
- H04N21/25891—Management of end-user data being end-user preferences
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Computer Graphics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention belongs to the technical field of information push, and particularly provides an information push strategy generation system and an information push strategy generation method, wherein the system comprises the following steps: the storage module is used for storing behavior data of a user; the calling module is used for calling the behavior data of the user in the storage module for nearly x days; the processing module is used for calculating the influence value of each content tag on the user according to the behavior data of the user called by the calling module and finding the preference tag of the user; the grouping module is used for carrying out preference grouping on the users according to the preference labels of the users; the matching module is used for matching the keywords of the pushed content with the tags grouped by the users; the pushing module is used for pushing the pushing content to the matched user group; the storage module is also used for storing the users according to the preference groups. By the method and the device, the content can be pushed to the user more accurately.
Description
Technical Field
The invention belongs to the technical field of information pushing, and particularly relates to an information pushing strategy generating system and method.
Background
With the development of interconnection technology, various aspects of people's life have changed greatly. The rise of modes such as internet of things, internet +, O2O and the like thoroughly changes the living habits of people.
IPTV, it is a technological product that appears in such a background. The IPTV is a brand-new technology that integrates internet, multimedia, communication and other technologies, and provides a variety of interactive services including digital tv to home users.
Due to the flexibility and variety of IPTV functions, the viscosity of its users is high. It is becoming more and more important to know the usage habits of users and to push the corresponding content directionally.
After the using habits of the users are known, the content business can be adjusted according to the using habits of the users, the user viscosity is further increased, decision support can be provided for advertisement putting push of operators, the effect of putting advertisements is better, advertising and recruitment are facilitated, and the profit is increased.
Currently, when IPTV is used to deliver content, a non-directional delivery method is mostly used, that is, delivered content is inserted in a fixed column or a fixed time period. Even if the personal attributes of the user are considered in content delivery, the system only extracts the program types frequently seen by the user and then directionally delivers the content related to the program types.
Although the number of users of the IPTV is large, the delivery method can achieve a certain delivery effect. However, the current interest point of the user is not fully considered in such a delivery mode, and when the user receives the delivered content, the user is likely to have little interest in the content or even no interest in the content.
Disclosure of Invention
The invention provides an information push strategy generation system and method aiming at the problem that the current interest point of a user is not fully considered when content is delivered in the prior art.
The basic scheme provided by the invention is as follows:
an information push policy generation system comprising:
the storage module is used for storing behavior data of a user;
the calling module is used for calling the behavior data of the user in the storage module for nearly x days;
the processing module is used for calculating the influence value of each content tag on the user according to the behavior data of the user called by the calling module and finding the preference tag of the user;
the grouping module is used for carrying out preference grouping on the users according to the preference labels of the users;
the matching module is used for matching the keywords of the pushed content with the tags grouped by the users;
the pushing module is used for pushing the pushing content to the matched user group;
the storage module is also used for storing the users according to the preference groups.
The noun explains: content tags, i.e., tags of column types of IPTV, such as genre, art, city, legal, etc.;
preference tags, i.e., content tags preferred by the user.
Basic scheme theory of operation and beneficial effect:
after the storage module stores the behavior data of the user, the calling module calls the behavior data of the user in the last X days, and the specific numerical value of X can be specifically set by a person skilled in the art according to a specific timeline standard; then, the processing module calculates the influence value of each content label on the user according to the called behavior data of the user, and finds out the preference label of the user; the grouping module groups the users according to the preference labels; when the content needs to be pushed, the matching module matches the keywords of the pushed content with the tags of the user groups to find the audience groups of the pushed content, and finally, the pushing module pushes the pushed content to the matched user groups.
When the method and the device are used for grouping the users, the interests of the users are considered, the influence degree (influence value) of each interest on the users is also considered comprehensively, so that the obtained user grouping can more reasonably reflect the current interest preference of the users, and compared with the prior art, the method and the device can more accurately push the contents of the users.
Further, the processing module comprises a screening submodule, a classification submodule, a base submodule, a weighting submodule, an accumulation submodule, a sequencing submodule and a marking submodule;
the screening submodule is used for screening out effective behavior data of the user and screening out playing behaviors of which the user stay time is longer than preset time t 1;
the classification submodule is used for classifying the screened behavior data of the user according to a preset content label;
the cardinality submodule is used for carrying out cardinality value judgment on the classified behavior data, the cardinality value of the behavior data which is larger than t1 and smaller than t2 is judged as y1, the cardinality value of the behavior data which is larger than t2 and smaller than t3 is judged as y2, and the cardinality value of the behavior data which is larger than t3 is judged as y3, wherein t3> t2> t1, y3> y2> y 1;
the weighting submodule is used for calculating the weight of the behavior data by using a preset weight calculation formula according to the distance between the screened behavior data and the current time, and weighting the behavior data of the user after the base number value is judged;
the accumulation submodule is used for accumulating the weighted behavior data of the user in each content tag to obtain an influence value of each content tag;
the sequencing submodule is used for sequencing the influence value of each content label by using a sequencing algorithm;
the marking submodule is used for marking the user by taking the content label of the P before the ranking in the sequencing submodule as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags.
Has the advantages that:
the screening submodule screens effective behavior data larger than t1 in the behavior data of the user, and the classification submodule classifies the screened behavior data of the user according to the content tags, so that the behavior data in each content tag can be conveniently and intensively processed.
Then, the cardinality sub-module performs basic value determination on the classified behavior data, and specific values of t1, t2, t3, y1, y2 and y3 can be specifically set by those skilled in the art according to specific situations. The longer the continuous viewing time is, the stronger the interest degree of the user in the type of the column is, and the base sub-module can roughly obtain the attention degree of the user represented behind each behavior data to the content label. Although the program may have to be stopped watching because of work and life, these are rare situations, and the probability of occurrence in each content label is the same, which does not affect the accuracy of the whole data.
And then, the weighting submodule calculates the weight of the behavior data according to a preset weight calculation formula, and weights the behavior data of the user after the base number value is judged. Most people have differences in the emphasis of interest over different time periods. For example, a user who prefers sports and history, the focus of the week is sports (e.g., a certain league he is interested in enters the stage of blanching), and the focus of the next week may be shifted to history (e.g., a certain student he likes develops a series of lectures).
Therefore, the interest tags need to be weighted, and the interest tags farther away from the current place have smaller influence on the user and should be weakened, so that the interest tags need to be weighted and the value of the weight value becomes smaller with the passage of time. Interest tags that are further away in time can be weakened by the weighting sub-module. The content which the user is interested in can be pushed more accurately.
After the sequencing submodule sequences the influence values of all the content labels by using a sequencing algorithm, the marking submodule is used for marking the user by taking the content label P before the sequencing in the sequencing submodule as a preference label of the user, so that P column types with the highest preference priority of the user are screened out. The specific value of P can be specifically set by those skilled in the art according to specific situations.
Through the sub-modules, the processing module can accurately find the content tag which has the largest influence on the user at present as the preference tag of the user according to the behavior data of the user.
Furthermore, in the weighting submodule, the preset weight calculation formula is a user-defined time forgetting function
Wherein t is the time from the current time.
Has the advantages that:
the weighted value calculation formula is a self-defined forgetting simulation function
And fitting the curve with an Ebinghaos forgetting curve, wherein n is a forgetting coefficient which is more than 0 and less than 1, a represents a short-term memory forgetting rate, and B represents a long-term forgetting rate. Through system simulation, when n is 0.25, a is 0.42 and b is 0.0003, the time decay function is most closely fitted to the Ebinghaos forgetting curve.
That is to say that the first and second electrodes,
the degree of fitting with an Ebinghaos forgetting curve is very high, and the influence degree of the interest label on the user can be well represented.
Further, the sorting algorithm adopted by the sorting submodule is a bubble sorting algorithm, an insertion sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm or a merging sorting algorithm.
Has the advantages that:
compared with the Hill sorting algorithm, the selective sorting algorithm, the heap sorting algorithm and the rapid sorting algorithm, the sorting algorithms have stronger stability, can continuously and stably carry out sorting work, and can prevent a disordered sorting result.
The system further comprises an updating module, a judging module and a judging module, wherein the updating module is used for updating the user members of the preference group in the storage module; and the EPG page updating module is also used for updating the EPG page of the user according to the preference tag of the user.
Has the advantages that:
the preference of the user changes along with the lapse of time, and the preference groups can be updated in time according to the change of the preference of the user through the updating module; meanwhile, the updating module also adjusts the EPG page of the user in time by combining the preference change of the user, so that the user can conveniently and quickly find the favorite column type, and the use experience of the user can be enhanced.
Further, the system also comprises a modification module used for modifying the content label.
Has the advantages that:
since the program type of the IPTV can be continuously adjusted along with the operation effect, after a long time, for example, after 1 year, the frame of the program type may have been changed greatly, and through the modification module, the manager can modify the content tag in time, so that the content tag can be always consistent with the current program type frame.
Based on the above system, the present application further provides an information pushing policy generating method, including:
a storage step of storing behavior data of a user;
calling, namely calling the behavior data of the user in the storage module for nearly x days;
a processing step, namely calculating the influence value of each content tag on the user according to the behavior data of the user called in the calling step, and finding out the preference tag of the user;
grouping, namely performing preference grouping on the users according to preference labels of the users;
a storage step, storing the users according to the preference grouping;
matching, namely matching the keywords of the pushed content with the tags grouped by the users;
and pushing the pushed content to the matched user group.
Has the advantages that:
when the users are grouped, the interests of the users are considered, and the influence degrees (influence values) of the interests on the users are also comprehensively considered, so that the obtained user groups can more reasonably reflect the current interest preferences of the users, and compared with the prior art, the content push can be more accurately carried out on the users.
Further, the processing step includes a screening sub-step, a classification sub-step, a cardinality sub-step, a weighting sub-step, an accumulation sub-step, a sorting sub-step, and a marking sub-step;
screening effective behavior data of the user, and screening the playing behavior of which the user dwell time is longer than a preset time t 1;
a classification substep, classifying the screened behavior data of the user according to a preset content label;
a cardinality substep of performing cardinality value judgment on the classified behavior data, judging the cardinality value of the behavior data which is larger than t1 and smaller than t2 as y1, judging the cardinality value of the behavior data which is larger than t2 and smaller than t3 as y2, and judging the cardinality value of the behavior data which is larger than t3 as y3, wherein t3> t2> t1, y3> y2> y 1;
a weighting substep, calculating the weight of the behavior data by using a preset weight calculation formula according to the time from the screened behavior data to the current time, and weighting the behavior data of the user after the base number value judgment;
an accumulation substep, accumulating the weighted behavior data of the user in each content label to obtain the influence value of each content label;
a sorting substep, sorting the influence values of the content labels by using a sorting algorithm;
marking, namely marking the content label of the P before the ranking in the sorting substep as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags.
Has the advantages that:
through the sub-steps, the processing step can accurately find the content label which has the largest influence on the user at present as the preference label of the user according to the behavior data of the user.
Further, in the weighting substep, the preset weight calculation formula is a user-defined time forgetting function
Wherein t is the time from the current time.
Has the advantages that:
the degree of fitting with an Ebinghaos forgetting curve is very high, and the influence degree of the interest label on the user can be well represented.
Further, in the sorting substep, the sorting algorithm adopted is a bubble sorting algorithm, an insert sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm or a merge sorting algorithm.
Has the advantages that:
compared with the Hill sorting algorithm, the selective sorting algorithm, the heap sorting algorithm and the rapid sorting algorithm, the sorting algorithms have stronger stability, can continuously and stably carry out sorting work, and can prevent a disordered sorting result.
Drawings
Fig. 1 is a logic block diagram of a first embodiment of an information push policy generation system according to the present invention;
FIG. 2 is a logical block diagram of the processing module of FIG. 1;
fig. 3 is a flowchart of a first embodiment of a method for generating an information push policy according to the present invention;
fig. 4 is a flow chart of the processing steps in fig. 3.
Detailed Description
The following is further detailed by way of specific embodiments:
as shown in fig. 1:
an information push strategy generation system comprises a server and a management end. The management terminal communicates with the server through the prior art, such as a WIFI module.
Server
In this embodiment, the server is a cloud server, and in other embodiments, the server may also be a Tencent cloud server or a distributed server.
The server comprises a storage module, a calling module, a processing module, a grouping module, an updating module, a matching module and a pushing module.
The storage module is used for storing behavior data of a user; and also for storing the users in groups of preferences.
The calling module is used for calling the behavior data of the user in the storage module for nearly x days; in this embodiment, x has a value of 30. If the time line is too short, the behavior of the user cannot be objectively reflected; if the time line is too long, the preference of the user is likely to change greatly, and the reference significance of the former behavior data is small; the behavior data of the near 30 days can better reflect the recent behaviors of the user on the basis of higher referential property.
And the processing module is used for calculating the influence value of each content tag on the user according to the behavior data of the user called by the calling module and finding the preference tag of the user.
The processing module updates the preference tag every m days, and in this embodiment, the value of m is 3. The preference label of the user cannot change every day, and meanwhile, the preference label is updated once a day, so that the load on the system is large; however, if the preference tag is updated only once a week, the preference of the user cannot be tracked in time when the preference of the user is transferred; the preference label is updated once in 3 days, so that too large load can not be caused to the system, and meanwhile, the preference label of the user can be updated in time according to the interest trend of the user.
Specifically, as shown in fig. 2, the processing module includes a screening sub-module, a classification sub-module, a cardinality sub-module, a weighting sub-module, an accumulation sub-module, a sorting sub-module, and a marking sub-module.
The screening submodule is used for screening out effective behavior data of the user and screening out playing behaviors of which the user stay time is longer than preset time t 1;
and the classification submodule is used for classifying the screened behavior data of the user according to a preset content label. The content tag is the column type of the content watched by the user, such as genre, art, city, legal, and so on. The names and the number of the content tags can be specifically set by those skilled in the art according to the specific column architecture of the IPTV.
And the cardinality submodule is used for carrying out cardinality value judgment on the classified behavior data, judging the cardinality value of the behavior data which is larger than t1 and smaller than t2 as y1, judging the cardinality value of the behavior data which is larger than t2 and smaller than t3 as y2, and judging the cardinality value of the behavior data which is larger than t3 as y3, wherein t3> t2> t1, and y3> y2> y 1. the specific values of t1, t2, t3, y1, y2 and y3 can be set by those skilled in the art according to specific situations. In this example, t1 is 5 minutes, t2 is 20 minutes, t3 is 45 minutes, y1 is 1, y2 is 1.5, and y3 is 2.
The longer the continuous viewing time, the more strongly the user's interest in the type of column is indicated. Although it is possible that the program has to be stopped watching because of work and life, these are rare situations, and the probability of occurrence in various content labels is the same, and the accuracy of the whole data is not affected.
The weighting submodule is used for calculating the weight of the behavior data by using a preset weight calculation formula according to the distance between the screened behavior data and the current time, and weighting the behavior data of the user after the base number value is judged;
the accumulation submodule is used for accumulating the weighted behavior data of the user in each content tag to obtain an influence value of each content tag;
and the sequencing submodule is used for sequencing the influence value of each content label by using a sequencing algorithm. In the embodiment, the sorting algorithm is a bubble sorting algorithm, and compared with other sorting algorithms, the bubble sorting algorithm has the advantages of simple related program, low spatial complexity, very stable operation result and capability of continuously and stably sorting. In other embodiments, an insert sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm or a merge sorting algorithm can be adopted, and the stability of the algorithms is also excellent.
The marking submodule is used for marking the user by taking the content label of the P before the ranking in the sequencing submodule as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags. The specific value of P can be set by those skilled in the art according to the strictness of the precision preferred by the user, and in this embodiment, the value of P is 3. Typically, the preferences of a user are not too broad, and 3 types of content tags are sufficient to cover the viewing preferences of most users.
In this embodiment, the weight calculation formula preset in the weighting submodule is a user-defined time forgetting function
The weighted value calculation formula is a self-defined forgetting simulation function
And fitting the curve with an Ebinghaos forgetting curve, wherein n is a forgetting coefficient which is more than 0 and less than 1, a represents a short-term memory forgetting rate, and B represents a long-term forgetting rate. Through system simulation, when n is 0.25, a is 0.42 and b is 0.0003, the time decay function is most closely fitted to the Ebinghaos forgetting curve.
That is to say that the first and second electrodes,the degree of fitting with an Ebinghaos forgetting curve is very high, and the influence degree of the interest label on the user can be well represented.
Most people have varying emphasis on interest at different times. For example, a user who prefers sports and history, the focus of the week is sports (e.g., a certain league he is interested in enters the stage of blanching), and the focus of the next week may be shifted to history (e.g., a certain student he likes develops a series of lectures). Therefore, the interest tags need to be weighted, and the interest tags farther away from the current place have smaller influence on the user and should be weakened, so that the interest tags need to be weighted and the value of the weight value becomes smaller with the passage of time.
The user-defined time decay function is used as the weight value to weight the interest tags, then the content tags are sequenced, the ranking list of the obtained content tags can reflect the degree of influence of the content tags on the user more truly, and managers can push the content more consistent with the current interest point of the user to the user.
The grouping module is used for grouping the preference of the users according to the preference labels of the users.
The updating module is used for updating the preference group in the storage module; the updating module is also used for updating the EPG page of the user according to the preference tag of the user.
The preference of the user changes along with the lapse of time, and the preference groups can be updated in time according to the change of the preference of the user through the updating module; meanwhile, the updating module also adjusts the EPG page of the user in time by combining the preference change of the user, so that the user can conveniently and quickly find the favorite column type, and the use experience of the user can be enhanced.
And the matching module is used for matching the keywords of the pushed content with the tags grouped by the users.
And the pushing module is used for pushing the pushed content to the matched user group.
Management terminal
In this embodiment, the management terminal is a PC.
The management terminal comprises a viewing module, a grouping calling module and a modifying module.
The viewing module is used for viewing the grouped data in the storage module of the server. Through the view module, the administrator can know the number of users in each group.
The packet calling module is used for calling the packet data in the storage module. Through the grouping calling module, the pushed content can be accurately delivered. If the user cooperates with a client of a certain sports goods to carry out advertisement putting, the preference label can be called to carry out advertisement putting for the users of the genre.
The modification module is used for modifying the content label. Since the program type of the IPTV can be continuously adjusted along with the operation effect, after a long time, for example, after 1 year, the frame of the program type may have been changed greatly, and through the modification module, the manager can modify the content tag in time, so that the content tag can be always consistent with the current program type frame.
Based on the above system, as shown in fig. 3, the present application further provides an information pushing policy generating method, including:
a storage step of storing behavior data of a user;
calling, namely calling the behavior data of the user in the storage module for nearly x days; the value of x in this example is 30.
A processing step, namely calculating the influence value of each content tag on the user according to the behavior data of the user called in the calling step, and finding out the preference tag of the user;
grouping, namely performing preference grouping on the users according to preference labels of the users;
a storage step, storing the users according to the preference grouping;
matching, namely matching the keywords of the pushed content with the tags grouped by the users;
pushing, namely pushing the pushed content to the matched user group;
and updating the user members of the preference group, and/or updating the EPG page of the user according to the preference tag of the user.
And a checking step, namely checking the grouped data in the storage module.
And a packet calling step of calling the packet data in the storage module.
And a modification step, namely modifying the content label.
Wherein, as shown in FIG. 4, the processing step includes a screening sub-step, a classification sub-step, a cardinality sub-step, a weighting sub-step, an accumulation sub-step, a sorting sub-step, and a marking sub-step;
screening effective behavior data of the user, and screening the playing behavior of which the user dwell time is longer than a preset time t 1;
a classification substep, classifying the screened behavior data of the user according to a preset content label;
a cardinality substep of performing cardinality value judgment on the classified behavior data, judging the cardinality value of the behavior data which is larger than t1 and smaller than t2 as y1, judging the cardinality value of the behavior data which is larger than t2 and smaller than t3 as y2, and judging the cardinality value of the behavior data which is larger than t3 as y3, wherein t3> t2> t1, y3> y2> y 1; in this example, t1 is 5 minutes, t2 is 20 minutes, t3 is 45 minutes, y1 is 1, y2 is 1.5, and y3 is 2.
A weighting substep, calculating the weight of the behavior data by using a preset weight calculation formula according to the time from the screened behavior data to the current time, and weighting the behavior data of the user after the base number value judgment; in this embodiment, the preset weight calculation formula is a custom time forgetting function
Wherein t is the time from the current time;
an accumulation substep, accumulating the weighted behavior data of the user in each content label to obtain the influence value of each content label;
a sorting substep, sorting the influence values of the content labels by using a sorting algorithm; in this embodiment, the sorting algorithm used is a bubble sorting algorithm, and in other embodiments, an insertion sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm, or a merge sorting algorithm may also be used, which are excellent in stability.
Marking, namely marking the content label of the P before the ranking in the sorting substep as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags.
Example two
Different from the first embodiment, the system in this embodiment further includes a user side, where the user side includes a user information storage module and a user analysis module;
the user information storage module is used for storing an information table of the user, in this embodiment, the information table of the user includes an age, a gender, a mobile phone ID and a preference tag of each user, and the preference tag in the user information table and the preference tag in the server are updated synchronously.
In this embodiment, the storage module stores behavior data of the user according to the mobile phone ID of the user, and the processing module analyzes the preference tag of the mobile phone ID of each user.
The user analysis module is used for analyzing the user watching the program at present according to the flow using condition of the mobile phone ID; specifically, when it is detected that the mobile phone ID of a certain user is accessed to WIFI, but the traffic usage of the mobile phone ID of the user within K seconds is less than L, it is determined that the user is watching a program; the user analysis module is also used for extracting the information of the user watching the program and sending the information to the server, and the server pushes the program according to the preference label of the user watching the program. The specific values of K and L can be set by those skilled in the art according to the severity of the user's selection.
There are typically multiple members in a family, and the preferred program types will vary from family member to family member due to gender, age, and hobbies. In this embodiment, each user has its own preference tag, and after knowing a specific user watching a program, the user analysis module sends the preference tag of the user to the server, and the server pushes the corresponding program according to the preference tag of the user.
Thus, when the user turns on the television for watching, the program which is interested by the user can be found out in the first time.
When the user watching the program has only one bit and changes, the system can timely find and carry out corresponding program pushing. The newly watched users can find the programs which are interested in themselves in the shortest time.
When the user who originally watches the program still watches but adds a new watching user, the system can push the new program, and each user can select the corresponding program to watch after negotiation and unified idea. Through such a mode, given each user more watching the selection, simultaneously, also given each user and deepened the chance of understanding each other between, the life rhythm of current society is very fast, and a family is in the same place the time also less and less, through the propelling movement of program, can know other people's recent focus between the family member, and then deepens each other's understanding, is convenient for find common topic.
The foregoing is merely an example of the present invention, and common general knowledge in the field of known specific structures and characteristics is not described herein in any greater extent than that known in the art at the filing date or prior to the priority date of the application, so that those skilled in the art can now appreciate that all of the above-described techniques in this field and have the ability to apply routine experimentation before this date can be combined with one or more of the present teachings to complete and implement the present invention, and that certain typical known structures or known methods do not pose any impediments to the implementation of the present invention by those skilled in the art. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.
Claims (10)
1. An information push policy generation system, comprising:
the storage module is used for storing behavior data of a user;
the calling module is used for calling the behavior data of the user in the storage module for nearly x days;
the processing module is used for calculating the influence value of each content tag on the user according to the behavior data of the user called by the calling module and finding the preference tag of the user;
the grouping module is used for carrying out preference grouping on the users according to the preference labels of the users;
the matching module is used for matching the keywords of the pushed content with the tags grouped by the users;
the pushing module is used for pushing the pushing content to the matched user group;
the storage module is also used for storing the users according to the preference groups.
2. The information push policy generation system according to claim 1, wherein: the processing module comprises a screening submodule, a classification submodule, a base submodule, a weighting submodule, an accumulation submodule, a sorting submodule and a marking submodule;
the screening submodule is used for screening out effective behavior data of the user and screening out playing behaviors of which the user stay time is longer than preset time t 1;
the classification submodule is used for classifying the screened behavior data of the user according to a preset content label;
the cardinality submodule is used for carrying out cardinality value judgment on the classified behavior data, the cardinality value of the behavior data which is larger than t1 and smaller than t2 is judged as y1, the cardinality value of the behavior data which is larger than t2 and smaller than t3 is judged as y2, and the cardinality value of the behavior data which is larger than t3 is judged as y3, wherein t3> t2> t1, y3> y2> y 1;
the weighting submodule is used for calculating the weight of the behavior data by using a preset weight calculation formula according to the distance between the screened behavior data and the current time, and weighting the behavior data of the user after the base number value is judged;
the accumulation submodule is used for accumulating the weighted behavior data of the user in each content tag to obtain an influence value of each content tag;
the sequencing submodule is used for sequencing the influence value of each content label by using a sequencing algorithm;
the marking submodule is used for marking the user by taking the content label of the P before the ranking in the sequencing submodule as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags.
3. The information push policy generation system according to claim 2, wherein: in the weighting submodule, the preset weight calculation formula is a user-defined time forgetting function
Wherein t is the time from the current time.
4. The information push policy generation system according to claim 2, wherein: the sorting algorithm adopted by the sorting submodule is a bubble sorting algorithm, an insertion sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm or a merging sorting algorithm.
5. The information push policy generation system according to claim 1, wherein: the updating module is used for updating the user members of the preference group in the storage module; and the EPG page updating module is also used for updating the EPG page of the user according to the preference tag of the user.
6. The information push policy generation system according to claim 1, wherein: the system also comprises a modification module used for modifying the content label.
7. An information push strategy generation method is characterized by comprising the following steps:
a storage step of storing behavior data of a user;
calling, namely calling the behavior data of the user in the storage module for nearly x days;
a processing step, namely calculating the influence value of each content tag on the user according to the behavior data of the user called in the calling step, and finding out the preference tag of the user;
grouping, namely performing preference grouping on the users according to preference labels of the users;
a storage step, storing the users according to the preference grouping;
matching, namely matching the keywords of the pushed content with the tags grouped by the users;
and pushing the pushed content to the matched user group.
8. The information push policy generation method according to claim 7, wherein: the processing step comprises a screening sub-step, a classification sub-step, a cardinality sub-step, a weighting sub-step, an accumulation sub-step, a sorting sub-step and a marking sub-step;
screening effective behavior data of the user, and screening the playing behavior of which the user dwell time is longer than a preset time t 1;
a classification substep, classifying the screened behavior data of the user according to a preset content label;
a cardinality substep of performing cardinality value judgment on the classified behavior data, judging the cardinality value of the behavior data which is larger than t1 and smaller than t2 as y1, judging the cardinality value of the behavior data which is larger than t2 and smaller than t3 as y2, and judging the cardinality value of the behavior data which is larger than t3 as y3, wherein t3> t2> t1, y3> y2> y 1;
a weighting substep, calculating the weight of the behavior data by using a preset weight calculation formula according to the time from the screened behavior data to the current time, and weighting the behavior data of the user after the base number value judgment;
an accumulation substep, accumulating the weighted behavior data of the user in each content label to obtain the influence value of each content label;
a sorting substep, sorting the influence values of the content labels by using a sorting algorithm;
marking, namely marking the content label of the P before the ranking in the sorting substep as a preference label of the user; when the number of the sequenced content tags is less than P, all the content tags are used as preference tags.
9. The information push policy generation method according to claim 8, wherein: in the weighting substep, the preset weight calculation formula is a user-defined time forgetting function
Wherein t is the time from the current time.
10. The information push policy generation method according to claim 8, wherein: in the sorting substep, the sorting algorithm adopted is a bubble sorting algorithm, an insert sorting algorithm, a bucket sorting algorithm, a radix sorting algorithm or a merge sorting algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910994371.9A CN110753247B (en) | 2019-10-18 | 2019-10-18 | Information push strategy generation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910994371.9A CN110753247B (en) | 2019-10-18 | 2019-10-18 | Information push strategy generation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110753247A true CN110753247A (en) | 2020-02-04 |
| CN110753247B CN110753247B (en) | 2021-11-09 |
Family
ID=69278893
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910994371.9A Active CN110753247B (en) | 2019-10-18 | 2019-10-18 | Information push strategy generation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110753247B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111324733A (en) * | 2020-02-07 | 2020-06-23 | 北京创鑫旅程网络技术有限公司 | Content recommendation method, device, equipment and storage medium |
| CN111683154A (en) * | 2020-06-17 | 2020-09-18 | 腾讯科技（深圳）有限公司 | Content pushing method, device, medium and electronic equipment |
| CN112364261A (en) * | 2020-11-30 | 2021-02-12 | 重庆空间视创科技有限公司 | Multimodal intelligent social information pushing system and method |
| WO2021163856A1 (en) * | 2020-02-17 | 2021-08-26 | 深圳市欢太科技有限公司 | Content pushing method and apparatus, and server and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106028071A (en) * | 2016-05-17 | 2016-10-12 | Tcl集团股份有限公司 | Video recommendation method and system |
| CN106980662A (en) * | 2017-03-21 | 2017-07-25 | 上海星红桉数据科技有限公司 | Based on magnanimity across the user tag sorting technique for shielding viewing behavior data |
| CN107222566A (en) * | 2017-07-06 | 2017-09-29 | 广州优视网络科技有限公司 | Information-pushing method, device and server |
-
2019
- 2019-10-18 CN CN201910994371.9A patent/CN110753247B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106028071A (en) * | 2016-05-17 | 2016-10-12 | Tcl集团股份有限公司 | Video recommendation method and system |
| CN106980662A (en) * | 2017-03-21 | 2017-07-25 | 上海星红桉数据科技有限公司 | Based on magnanimity across the user tag sorting technique for shielding viewing behavior data |
| CN107222566A (en) * | 2017-07-06 | 2017-09-29 | 广州优视网络科技有限公司 | Information-pushing method, device and server |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111324733A (en) * | 2020-02-07 | 2020-06-23 | 北京创鑫旅程网络技术有限公司 | Content recommendation method, device, equipment and storage medium |
| WO2021163856A1 (en) * | 2020-02-17 | 2021-08-26 | 深圳市欢太科技有限公司 | Content pushing method and apparatus, and server and storage medium |
| CN111683154A (en) * | 2020-06-17 | 2020-09-18 | 腾讯科技（深圳）有限公司 | Content pushing method, device, medium and electronic equipment |
| CN111683154B (en) * | 2020-06-17 | 2023-11-14 | 腾讯科技（深圳）有限公司 | Content pushing method, device, medium and electronic equipment |
| CN112364261A (en) * | 2020-11-30 | 2021-02-12 | 重庆空间视创科技有限公司 | Multimodal intelligent social information pushing system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110753247B (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110753247B (en) | Information push strategy generation system and method | |
| US11409792B2 (en) | System and method for streaming individualized media content | |
| US8930294B2 (en) | Predicting user activity based on usage data received from client devices | |
| CN108028962B (en) | Processing video usage information to deliver advertisements | |
| US8732017B2 (en) | Methods, systems, and media for applying scores and ratings to web pages, web sites, and content for safe and effective online advertising | |
| CN110795624A (en) | Analyzing a personalization framework | |
| KR101060487B1 (en) | Apparatus and method for content recommendation using tag cloud | |
| US20140379746A1 (en) | System and method for providing topic cluster based updates | |
| US20120078725A1 (en) | Method and system for contextual advertisement recommendation across multiple devices of content delivery | |
| US20110196927A1 (en) | Social Networking Application Using Posts to Determine Compatibility | |
| US20100228591A1 (en) | Real time ad selection for requested content | |
| CN106131703A (en) | A kind of method of video recommendations and terminal | |
| CN111680254B (en) | Content recommendation method and device | |
| CN110474944B (en) | Network information processing method, device and storage medium | |
| US11727221B2 (en) | Dynamic correlated topic model | |
| CN113742567A (en) | Multimedia resource recommendation method and device, electronic equipment and storage medium | |
| CN114339417A (en) | Video recommendation method, terminal device and readable storage medium | |
| US20220414716A1 (en) | Managing metadata enrichment of digital asset portfolios | |
| CN111090810A (en) | Application message pushing method and device and electronic equipment | |
| CN110753248B (en) | System and method for collecting IPTV user terminal data | |
| CN109756759B (en) | Bullet screen information recommendation method and device | |
| WO2010034056A1 (en) | Advertisement selection for a portable computing device | |
| EP2083385A1 (en) | Method of adapting a user profile including user preferences and communication device | |
| US20240005360A1 (en) | Multimedia enrichment and rendering with computing devices | |
| CN117651165B (en) | Video recommendation method and device based on client data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: 401121 room 701, room 1, floor 7, building 11, No. 106, west section of Jinkai Avenue, Yubei District, Chongqing Patentee after: Space Shichuang (Chongqing) Technology Co.,Ltd. Address before: 401121 17-4, building 2, No. 70, middle section of Huangshan Avenue, Yubei District, Chongqing Patentee before: Chongqing Space Visual Creation Technology Co.,Ltd. |
|
| CP03 | Change of name, title or address |