CN109325179B - Content promotion method and device - Google Patents
Content promotion method and device Download PDFInfo
- Publication number
- CN109325179B CN109325179B CN201811081806.2A CN201811081806A CN109325179B CN 109325179 B CN109325179 B CN 109325179B CN 201811081806 A CN201811081806 A CN 201811081806A CN 109325179 B CN109325179 B CN 109325179B
- Authority
- CN
- China
- Prior art keywords
- user
- content
- weight
- tag
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Information Transfer Between Computers (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a device for content promotion, wherein the method comprises the following steps: the method comprises the steps of obtaining popularization content, extracting keywords of the popularization content, determining a user to be popularized according to the keywords of the popularization content and user labels in user portrait, wherein the user portrait comprises the user labels and weights corresponding to the user labels, the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by the user, and the popularization content is pushed to the user to be popularized. According to the scheme, the target users with accurate popularization content are obtained according to the dynamic behavior analysis of the websites browsed by the users and the similarity among the users, and the popularization accuracy is improved.
Description
Technical Field
The embodiment of the invention relates to the technical field of data mining, in particular to a content promotion method and device.
Background
In recent years, various applications and services on the internet have been increasing in a blowout manner, and in order to enable target users to quickly know and use the services, developers and service providers of the applications promote their applications and services through various approaches and platforms. How to enable developers to obtain target users as accurate as possible with the minimum popularization cost becomes a problem to be solved by various large Internet popularization platforms.
The current internet promotion platform generally adopts a pricing or bidding mode, and an application bid to be promoted is high enough to occupy a favorable advertisement space, so that the current internet promotion platform has a higher exposure rate, or regional content is promoted according to a specific region, so that the promotion pertinence is not enough.
Disclosure of Invention
The embodiment of the invention provides a content promotion method and device, which are used for improving the promotion accuracy rate by acquiring accurate target users.
The method for content promotion provided by the embodiment of the invention comprises the following steps:
acquiring popularization content, and extracting keywords of the popularization content;
determining a user to be promoted according to the keywords of the promotion content and the user tags in the user portrait; the user portrait comprises user labels and weights corresponding to the user labels, wherein the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by users;
and pushing the promotion content to the user to be promoted.
In the above embodiment, the weight corresponding to the user tag is determined according to the user identification weight, the time weight, the content weight and the behavior weight of the website browsed by the user, the user to be promoted is determined according to the keyword of the promotion content and the user tag in the user portrait, that is, the user portrait is formed by combining the time dimension information according to the dynamic behavior analysis of the user, and the promotion content is pushed accurately.
Optionally, the determining the weight corresponding to the user tag according to the user identification weight, the time weight, the content weight, and the behavior weight of the user browsing the website includes:
acquiring browsing content, user identification and browsing time of the website browsed by the user;
determining the user tag and a content weight corresponding to the user tag according to the browsing content;
determining the user identification weight corresponding to the user label according to the user identification and the browsing time corresponding to the user identification;
determining a time weight and a behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag;
and determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label.
In the above embodiment, the browsing content, the user identifier, and the browsing time of the user browsing the website are obtained, and the user identifier weight, the time weight, the content weight, and the behavior weight corresponding to the user tag are determined according to the browsing content, the user identifier, and the browsing time of the user browsing the website, so as to determine the weight corresponding to the user tag.
Optionally, the determining the user tag and the content weight corresponding to the user tag according to the browsing content includes:
determining keywords of the user tags according to the browsing content;
determining TF-IDF (Term Frequency-Inverse text Frequency) indexes corresponding to the keywords of the user tags according to the number of times of the keywords of the user tags appearing in the browsing content, the page number of browsing pages contained in the browsing content, the total number of words of each browsing page and the page number of the browsing pages containing the keywords of the user tags;
and determining the maximum TF-IDF index in the keywords of the plurality of user tags as the content weight corresponding to the user tag.
In the above embodiment, the keywords of the user tag are determined, and the maximum TF-IDF index in the keywords of the user tag is determined as the content weight corresponding to the user tag.
Optionally, the determining, according to the browsing time corresponding to the user tag, a time weight and a behavior weight corresponding to the user tag includes:
dividing a preset time period into a plurality of sub-time periods, and determining the browsing time in each sub-time period;
determining the time weight corresponding to the user label according to the starting time and the ending time of the user label browsed in each sub-period, the starting time and the ending time of each sub-period and the attenuation coefficient of each sub-period;
and determining the behavior weight corresponding to the user tag according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag.
In the above embodiment, an implementation manner of determining the time weight corresponding to the user tag is provided, and meanwhile, the behavior weight corresponding to the user tag is determined according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag. And measuring the corresponding behavior weight of the user label by adopting time weight. And preparing for subsequently determining the corresponding weight of the user label.
Optionally, the determining, according to the keyword of the promotion content and the user tag in the user portrait, a user to be promoted includes:
comparing the keywords of the promotion content with the user tags in the user images, and determining the matching degree of the promotion content and the user images;
and determining the user with the matching degree of the promotion content and the user picture higher than a matching threshold value as the user to be promoted.
In the above embodiment, the user whose matching degree between the promotion content and the user image is higher than the matching threshold is determined as the user to be promoted, that is, the user matched with the promotion content is determined.
Optionally, after the pushing the promotion content to the user to be promoted, the method further includes:
determining the similarity between the user portrait of the user to be promoted and the user portrait of each user according to the user tags in the user portrait and the weights corresponding to the user tags;
and pushing the promotion content to the user with the similarity of the user portrait of the user to be promoted larger than a first threshold value.
In the embodiment, the user with the similarity to the user portrait of the user to be promoted being greater than the first threshold value is determined, so that the target user of the promotion content can be avoided being omitted when the user to be promoted is determined according to the keywords of the promotion content and the user tags in the user portrait.
Optionally, before determining the user to be promoted according to the keyword of the promotion content and the user tag in the user representation, the method further includes:
and acquiring the initial popularization crowd category of the popularization content, and pushing the popularization content to users belonging to the initial popularization crowd category.
In the above embodiment, the promotion content is pushed to the users belonging to the preliminary promotion crowd category according to the preliminary promotion crowd category of the promotion content, and the promotion method is suitable for promoting the cold start stage.
Correspondingly, the embodiment of the invention also provides a device for content promotion, which comprises:
the system comprises an acquisition unit, a search unit and a search unit, wherein the acquisition unit is used for acquiring popularization content and extracting keywords of the popularization content;
the processing unit is used for determining a user to be promoted according to the keyword of the promotion content and the user label in the user portrait; the user portrait comprises user labels and weights corresponding to the user labels, wherein the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by users; and pushing the promotion content to the user to be promoted.
Optionally, the processing unit is specifically configured to:
acquiring browsing content, user identification and browsing time of the website browsed by the user;
determining the user tag and a content weight corresponding to the user tag according to the browsing content;
determining the user identification weight corresponding to the user label according to the user identification and the browsing time corresponding to the user identification;
determining a time weight and a behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag;
and determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label.
Optionally, the processing unit is specifically configured to:
determining keywords of the user tags according to the browsing content;
determining word frequency-inverse text frequency TF-IDF indexes corresponding to the keywords of the user tags according to the times of the keywords of the user tags appearing in the browsing content, the page number of browsing pages contained in the browsing content, the total vocabulary number of each browsing page and the page number of the browsing pages containing the keywords of the user tags;
and determining the maximum TF-IDF index in the keywords of the plurality of user tags as the content weight corresponding to the user tag.
Optionally, the processing unit is specifically configured to:
dividing a preset time period into a plurality of sub-time periods, and determining the browsing time in each sub-time period;
determining the time weight corresponding to the user label according to the starting time and the ending time of the user label browsed in each sub-period, the starting time and the ending time of each sub-period and the attenuation coefficient of each sub-period;
and determining the behavior weight corresponding to the user tag according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag.
Optionally, the processing unit is specifically configured to:
comparing the keywords of the promotion content with the user tags in the user images, and determining the matching degree of the promotion content and the user images;
and determining the user with the matching degree of the promotion content and the user picture higher than a matching threshold value as the user to be promoted.
Optionally, the processing unit is further configured to:
after the promotion content is pushed to the user to be promoted, determining the similarity between the user portrait of the user to be promoted and the user portrait of each user according to the user label in the user portrait and the weight corresponding to the user label;
and pushing the promotion content to the user with the similarity of the user portrait of the user to be promoted larger than a first threshold value.
Optionally, the processing unit is further configured to:
before determining a user to be promoted according to the keywords of the promotion content and the user label in the user portrait, acquiring a preliminary promotion crowd category of the promotion content, and pushing the promotion content to the user belonging to the preliminary promotion crowd category.
Correspondingly, an embodiment of the present invention further provides a computing device, including:
a memory for storing program instructions;
and the processor is used for calling the program instructions stored in the memory and executing the method popularized by the content according to the obtained program.
Accordingly, embodiments of the present invention further provide a computer-readable non-volatile storage medium, which includes computer-readable instructions, and when the computer-readable instructions are read and executed by a computer, the computer-readable instructions cause the computer to perform the method for content promotion.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic diagram of a system architecture according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a method for content promotion according to an embodiment of the present invention;
fig. 3 is a schematic flowchart of determining a weight corresponding to a user tag according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a content promotion apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 exemplarily shows a system architecture to which the method for providing content promotion according to the embodiment of the present invention is applicable, and the system architecture may include an advertisement access platform 101, a content promotion platform 102, and a user terminal 103. The advertisement access platform 101 may be an access platform for promotion content; the content promotion platform 102 is a platform for promoting content, which may be an internet platform, an APP, or the like, and the content promotion platform 102 is a comprehensive service platform that is wide in user object-oriented type and supports third-party access service or promotion of advertisements; the user terminal 103 may be a mobile phone, a tablet computer or other terminal that may promote advertising to the user.
Based on the above description, fig. 2 exemplarily shows a flow of a method for content promotion provided by the embodiment of the present invention, where the flow may be executed by a device for content promotion, and the device may be located in a content promotion platform, and may be the content promotion platform. As shown in fig. 2, the process specifically includes:
The promotion content is the content which needs to be promoted by an advertiser, and can be shopping advertisements, real news, entertainment news and the like. When the promotion content is obtained, the promotion content comprises the keywords and the function introduction of the promotion content, and preparation is made for determining the user to be promoted matched with the promotion content by extracting the keywords of the promotion content. For example, if the promotion content is a skin care product in a shopping advertisement, the keywords of the skin care product, such as moisturizing, acne removing and the like, can be extracted.
And step 202, determining a user to be promoted according to the keywords of the promotion content and the user tags in the user portrait.
The user portrait comprises a user label and a weight corresponding to the user label, wherein the weight corresponding to the user label is determined according to a user identification weight, a time weight, a content weight and a behavior weight of a website browsed by a user.
The user type and the user characteristics are analyzed by acquiring the entity static data and the dynamic interaction data of the website browsed by the user, and then the user matched with the popularization content is determined. The data used to create the user representation includes natural data, behavioral data, and content data. The natural data representation includes the user's own inherent attributes such as the user's gender and age, and can be collected from the user during the user registration and the like. The behavior data describes behaviors performed by the user, including access times, access frequency, access dwell time, operation active time, information input, link clicking, interactive operation (such as adding attention, removing attention, scoring, saving as a bookmark, adding a shopping cart, taking a shopping cart, forming an order, removing an order, paying, refunding and the like). The content data represents an object of a user behavior, such as a microblog ID (Identity) to which the user pays attention, a song scored by the user, content on a web page saved as a bookmark by the user, a commodity which the user adds to a shopping cart or forms an order, and the like.
As an implementation mode, based on data acquired by establishing user portrait, four factors of user identification, time, content and behavior of the user random internet behavior are analyzed by combining Spark task processing and Storm streaming processing, and the random internet behavior of the user can be comprehensively described. The weight corresponding to the user tag is determined according to the user identification weight, the time weight, the content weight and the behavior weight of the website browsed by the user, which may be specifically shown in fig. 3.
Here, the browsing website includes a plurality of browsing pages, the browsing pages include a plurality of keywords, and the keywords in the browsing pages may constitute the user tags.
Specifically, the keywords of the user tags are determined according to the browsing content. For example, if the user browsing content includes a keyword "mobile phone camera, mobile phone display chip, and mobile phone memory card", it may be determined that the keyword "mobile phone" labeled by the user is "mobile phone camera, mobile phone display chip, and mobile phone memory card".
After determining the keywords of the user tags, determining TF-IDF indexes corresponding to the keywords of the user tags according to the times of the keywords of the user tags appearing in the browsed contents, the number of pages of browsed pages contained in the browsed contents, the total number of words of each browsed page and the number of pages of the browsed pages containing the keywords of the user tags, and determining the maximum TF-IDF index in the keywords of the user tags as the content weight corresponding to the user tags.
Preferably, the content weight corresponding to the user tag may be as shown in formula (1).
The formula (1) is:
wherein, Ci(lj) For user UiUser tag ljCorresponding content weight, TF-IDF (l)j) A TF-IDF index corresponding to the keyword of the user tag, TF-IDF (l)j) As shown in equation (2).
The formula (2) is:
wherein, count (w) is the userThe number of times the keyword w of the tag appears in the viewed content,
total number of words for each viewed page, N total number of pages used for detection, NwAnd counting the number of pages containing the keyword w in all the N detected pages.
Here, the user identifier may be an identifier for distinguishing a user on the Internet, and the user identifier may be a Cookie, an IP (Internet Protocol, Protocol for interconnection between networks), Email, an identity card, or the like. Among other things, the Cookie value may be interpreted as data (typically encrypted) that certain websites store on the user's local terminal for purposes of user identity identification, session control tracking (session).
In this embodiment, the user identity is determined by the user behavior, the registration information, and the like. The user identification weight corresponding to the user label can be understood as the possibility that the user label is generated from the same user, and the total number of the users is assumed to be n, and the user is constructed to be U1，U2，……，UnUser tag ljGenerated from the same user UiMay be as shown in equation (3).
The formula (3) is:
wherein, Obji(lj) For user UiUser tag ljCorresponding subscriber identity weight,/jFor the jth user tag, the user tag,
for user UiThe accuracy of the mth identification of (1),
for user UiThe total browsing time of the mth identification of (1).
In addition, the longer the browsing time, the more available the user tags thereof. For example, if the browsing time of the user a is 10h and the browsing time of the user B is 1h, the user tag of the user a may be considered to be more available.
And step 304, determining the time weight and the behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag.
Specifically, the time includes a timestamp and a time interval, where the timestamp is used to identify the time when the event occurs and ends, and the time interval is used to identify the browsing time. Dividing a preset time period into a plurality of sub-time periods, determining the browsing time in each sub-time period, and determining the time weight corresponding to the user label according to the starting time and the ending time of the user label browsed in each sub-time period, the starting time and the ending time of each sub-time period and the attenuation coefficient of each sub-time period.
Preferably, the preset time period is divided into S sub-periods, T1，T2，…，Tk，…，TSWherein T iskIs the kth sub-period in the preset period, and each sub-period corresponds to a respective attenuation coefficient Ek。EkAs shown in equation (4).
The formula (4) is:
the lambda is a forgetting coefficient, the attenuation speed of the information along with time is determined by the value of the lambda, the attenuation speed is faster when the lambda is larger, and the attenuation speed is related to the number of the divided time periods. Lambda can be set to be 1-4 or more according to experience, and can be endowed with different values according to different user tag types. For example, the user label type with stronger timeliness has larger lambda assignment and faster decay speed.
The time weight corresponding to the user tag may be as shown in equation (5).
The equation (5) is:
wherein the content of the first and second substances,
for user UiUser tag l at kth sub-periodjThe corresponding weight of the time is given to,
browse user tab l for user in kth sub-periodjThe start time of the start,
browse user tab l for user in kth sub-periodjThe time of the termination of the operation of the mobile terminal,
the starting time of the k-th sub-period,
end time of kth sub-period, EkIs the attenuation coefficient of the kth sub-period.
The behavior of the user is the operation of the user on the website content, such as browsing, collecting, scoring, sharing and the like, the corresponding weights of different user behaviors are different, and the behavior weights generated on the user labels are also different. Specifically, the behavior weight corresponding to the user tag may be determined according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag.
When the operation of the user on the website content is browsing, the time weight corresponding to the user tag may be determined according to the browsing time corresponding to the user tag, which is also equivalent to the implementation of determining the time weight corresponding to the user tag in the foregoing embodiment.
When the operation of the user on the website content is collection, the collection weight corresponding to the user tag can be determined according to the formula (6).
The equation (6) is:
wherein the content of the first and second substances,
for user UiUser tag l at kth sub-periodjA corresponding collection weight;
to judge the function when
When it is determined
When in use
When it is determined
For user UiUser tag l at kth sub-periodjA corresponding time weight; alpha is a collection behavior coefficient corresponding to the user label, and alpha can be set to be more than or equal to 0.2 and less than or equal to 1 according to experience.
When the operation of the user on the website content is a score, the scoring weight corresponding to the user tag may be determined according to formula (7).
The equation (7) is:
wherein the content of the first and second substances,
for user UiUser tag l at kth sub-periodjCorresponding scoring weights;
is composed of
Is determined as a function of the average of (c),
for user UiUser tag l at kth sub-periodjA corresponding time weight;
is composed of
Is determined as a function of the average of (c),
for user UiFor user label ljScoring of (4); beta is a scoring behavior coefficient corresponding to the user label, and beta can be set to be more than or equal to 0.5 and less than or equal to 2 according to experience.
When the user operates the website content to share, the sharing weight corresponding to the user tag may be determined according to formula (8).
The equation (8) is:
wherein the content of the first and second substances,
for user UiUser tag l at kth sub-periodjCorresponding sharing weights;
is composed of
Is determined as a function of the average of (c),
for user UiUser tag l at kth sub-periodjA corresponding time weight; gamma is a sharing behavior coefficient corresponding to the user label, and can be set to be more than or equal to 0.2 and less than or equal to 3 according to experience.
It can be known from the above embodiments that the behavior weight corresponding to the user tag can be determined by the time weight corresponding to the user tag
And determining the behavior coefficients (collection behavior coefficient, scoring behavior coefficient and sharing behavior coefficient) corresponding to the user tags. Of course, it can also be said that the behavior weight corresponding to the user tag can be the time weight corresponding to the user tag
And (4) measuring.
And 305, determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label.
It can be said that the weight corresponding to the user tag can be determined by the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user tag. As one implementation, the weight corresponding to the user tag may be as shown in equation (9).
The formula (9) is:
wherein the content of the first and second substances,
for user UiUser tag ljCorresponding user tag weights; obji(lj) For user UiUser tag ljCorresponding subscriber identity weight, Ci(lj) For user UiUser tag ljThe corresponding weight of the content is set to,
for user UiUser tag l at kth sub-periodjThe corresponding weight of the time is given to,
for user UiUser tag l at kth sub-periodjA corresponding collection weight;
for user UiUser tag l at kth sub-periodjCorresponding scoring weights;
for user UiUser tag l at kth sub-periodjThe corresponding sharing weight. Wherein j is 1,2, … …, sumi。
Optionally, the user tags and the weights corresponding to the user tags are determined through the above embodiment, and then information aggregation of the user tags is performed based on the automatically mined user tags, so that tags with stronger readability and stronger descriptive property are obtained, and a user image is formed. For example, a user tag such as "mobile phone, tablet computer, VR device" automatically mined by a user is aggregated into a tag of "electronic product", and one user portrait may correspond to a plurality of aggregated tags.
And finally, determining the user to be promoted according to the keywords of the promotion content and the user label in the user portrait. Specifically, the keywords of the promotion content can be compared with the user tags in the user images to determine the matching degree of the promotion content and the user images, and then the user with the matching degree of the promotion content and the user images higher than the matching threshold value is determined as the user to be promoted. The user to be promoted here, that is, the user matched with the promotion content, may be a target user for pushing the promotion content.
Pushing the promotion content to the user to be promoted determined in step 202. Certainly, when the user to be promoted is determined and the promotion content is pushed to the user to be promoted, the keywords of the promotion content can be compared with the user tags in the user portrait, the matching degree of the promotion content and the user portrait is determined, the user portrait is sorted according to the matching degree, and the promotion content is preferentially pushed to the user with the matching degree sorted in front.
In addition, another way of determining a user to be promoted is provided in the embodiments of the present invention, after the promotion content is pushed to the user to be promoted, the similarity between the user portrait of the user to be promoted and the user portrait of each user can be determined according to the user tag in the user portrait and the weight corresponding to the user tag; and pushing the promotion content to the user with the similarity of the user portrait of the user to be promoted larger than a first threshold value. And equivalently, similar interest exists among users with similar user figures, and the users with similar user figures are determined according to the user figures of the users to be promoted. The like similarity between users can be calculated through a collaborative filtering correlation algorithm, and the co-browsing content is recommended to the users with similar like, and the specific implementation manner is as follows.
Firstly, calculating the similarity of user labels, combining the similar user labels, adding different user labels, and assigning the value of the user label of a user without the user label to be 0, thereby unifying the user portrait matrix LiRedefines user tag J to 1,2, …, J. User' s
The similarity therebetween may satisfy formula (10).
The equation (10) is:
wherein the content of the first and second substances,
for the user
The degree of similarity between the two images,
for the user
User tag l at kth sub-periodjThe corresponding user tag weight is set to be,
for the user
User tag l at kth sub-periodjCorresponding user tag weight.
By the method, the user with the similarity of the user portrait of the user to be promoted greater than the first threshold value is determined, and the target user of the promotion content can be prevented from being omitted when the user to be promoted is determined according to the keywords of the promotion content and the user tags in the user portrait.
For example, the user a is a user to promote the promotional content a, and the user a may like the promotional content a. When the user B is matched with the promotion content a, due to the fact that the matching degree of the determined promotion content and the user image is smaller than the matching threshold value due to the browsing time of the user B for browsing the website, the user identification and the like, the user B is judged not to be the user to be promoted of the promotion content a according to the matching degree, but the user image between the user B and the user A is very similar, the preference of the user B is similar to that of the user A, so that the user B can be determined to possibly like the promotion content a according to the user image similarity of the user B and the user A, and the promotion content a is pushed to the user B.
In addition, for the popularization content newly submitted on the content popularization platform, because user data is not accumulated yet, it is difficult to determine a user to be popularized, so the content popularization platform needs to provide an initial popularization object option for the advertisement access platform, that is, before determining the user to be popularized according to a keyword of the popularization content and a user tag in a user portrait, an initial popularization crowd category of the popularization content is obtained, and the popularization content is pushed to the user belonging to the initial popularization crowd category. For example, the preliminary promotional crowd categories may include: students, office workers, the old, women's white collar, fashion dawns and the like, if the popularization content is a pen, the preliminary popularization crowd category can be the students and the office workers; if the promotion content is lipstick, the primary promotion crowd category can be white collar of women and fashionable people.
In the above embodiment, the weight corresponding to the user tag is determined according to the user identification weight, the time weight, the content weight and the behavior weight of the website browsed by the user, the matching degree between the promotion content and the user image is determined according to the weight corresponding to the user tag and the user image, and finally the user matched with the promotion content is determined. Meanwhile, the promotion content is pushed to the users similar to the users to be promoted, and the defect of judgment through the matching degree of the promotion content and the user images is overcome. According to the analysis of the dynamic behaviors of the internet users, the user portrait is formed by combining time dimension information, and the contents on the internet platform are accurately popularized by combining the similarity between the users.
Based on the same inventive concept, fig. 4 exemplarily shows a structure of an apparatus for content promotion provided in an embodiment of the present invention, and the apparatus can execute a flow of a method for content promotion.
An obtaining unit 401, configured to obtain popularization content, and extract a keyword of the popularization content;
a processing unit 402, configured to determine a user to be promoted according to the keyword of the promotion content and the user tag in the user representation; the user portrait comprises user labels and weights corresponding to the user labels, wherein the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by users; and pushing the promotion content to the user to be promoted.
Optionally, the processing unit 402 is specifically configured to:
acquiring browsing content, user identification and browsing time of the website browsed by the user;
determining the user tag and a content weight corresponding to the user tag according to the browsing content;
determining the user identification weight corresponding to the user label according to the user identification and the browsing time corresponding to the user identification;
determining a time weight and a behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag;
and determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label.
Optionally, the processing unit 402 is specifically configured to:
determining keywords of the user tags according to the browsing content;
determining TF-IDF indexes corresponding to the keywords of the user tags according to the times of the keywords of the user tags appearing in the browsing content, the page number of browsing pages contained in the browsing content, the total vocabulary number of each browsing page and the page number of the browsing pages containing the keywords of the user tags;
and determining the maximum TF-IDF index in the keywords of the plurality of user tags as the content weight corresponding to the user tag.
Optionally, the processing unit 402 is specifically configured to:
dividing a preset time period into a plurality of sub-time periods, and determining the browsing time in each sub-time period;
determining the time weight corresponding to the user label according to the starting time and the ending time of the user label browsed in each sub-period, the starting time and the ending time of each sub-period and the attenuation coefficient of each sub-period;
and determining the behavior weight corresponding to the user tag according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag.
Optionally, the processing unit 402 is specifically configured to:
comparing the keywords of the promotion content with the user tags in the user images, and determining the matching degree of the promotion content and the user images;
and determining the user with the matching degree of the promotion content and the user picture higher than a matching threshold value as the user to be promoted.
Optionally, the processing unit 402 is further configured to:
after the promotion content is pushed to the user to be promoted, determining the similarity between the user portrait of the user to be promoted and the user portrait of each user according to the user label in the user portrait and the weight corresponding to the user label;
and pushing the promotion content to the user with the similarity of the user portrait of the user to be promoted larger than a first threshold value.
Optionally, the processing unit 402 is further configured to:
before determining a user to be promoted according to the keywords of the promotion content and the user label in the user portrait, acquiring a preliminary promotion crowd category of the promotion content, and pushing the promotion content to the user belonging to the preliminary promotion crowd category.
Based on the same inventive concept, an embodiment of the present invention further provides a computing device, including:
a memory for storing program instructions;
and the processor is used for calling the program instructions stored in the memory and executing the method popularized by the content according to the obtained program.
Based on the same inventive concept, the embodiment of the present invention further provides a computer-readable non-volatile storage medium, which includes computer-readable instructions, and when the computer reads and executes the computer-readable instructions, the computer-readable instructions cause the computer to execute the method for promoting the content.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (9)
1. A method of content promotion, comprising:
acquiring popularization content, and extracting keywords of the popularization content;
determining a user to be promoted according to the keywords of the promotion content and the user tags in the user portrait; the user portrait comprises user labels and weights corresponding to the user labels, wherein the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by users;
pushing the promotion content to the user to be promoted;
the determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight of the website browsed by the user comprises the following steps:
acquiring browsing content, user identification and browsing time of the website browsed by the user;
determining the user tag and a content weight corresponding to the user tag according to the browsing content;
determining the user identification weight corresponding to the user label according to the user identification and the browsing time corresponding to the user identification;
determining a time weight and a behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag;
determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label;
the weight corresponding to the user label can be shown as formula (9);
the formula (9) is:
wherein the content of the first and second substances,
for user UiUser tag ljCorresponding user tag weights; obji(lj) For user UiUser tag ljCorresponding subscriber identity weight, Ci(lj) For user UiUser tag ljThe corresponding weight of the content is set to,
for user UiUser tag l at kth sub-periodjThe corresponding weight of the time is given to,
for user UiUser tag l at kth sub-periodjA corresponding collection weight;
for user UiUser tag l at kth sub-periodjCorresponding scoring weights;
for user UiUser tag l at kth sub-periodjCorresponding sharing weights; wherein j is 1,2, … …, sumi。
2. The method of claim 1, wherein the determining the user tag and the content weight corresponding to the user tag according to the browsing content comprises:
determining keywords of the user tags according to the browsing content;
determining word frequency-inverse text frequency TF-IDF indexes corresponding to the keywords of the user tags according to the times of the keywords of the user tags appearing in the browsing content, the page number of browsing pages contained in the browsing content, the total vocabulary number of each browsing page and the page number of the browsing pages containing the keywords of the user tags;
and determining the maximum TF-IDF index in the keywords of the plurality of user tags as the content weight corresponding to the user tag.
3. The method of claim 1, wherein the determining the time weight and the behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag comprises:
dividing a preset time period into a plurality of sub-time periods, and determining the browsing time in each sub-time period;
determining the time weight corresponding to the user label according to the starting time and the ending time of the user label browsed in each sub-period, the starting time and the ending time of each sub-period and the attenuation coefficient of each sub-period;
and determining the behavior weight corresponding to the user tag according to the time weight corresponding to the user tag and the behavior coefficient corresponding to the user tag.
4. The method of claim 1, wherein determining the user to promote based on the keywords of the promotion content and the user tags in the user representation comprises:
comparing the keywords of the promotion content with the user tags in the user images, and determining the matching degree of the promotion content and the user images;
and determining the user with the matching degree of the promotion content and the user picture higher than a matching threshold value as the user to be promoted.
5. The method of claim 1, wherein after the pushing the promotional content to the user to promote, further comprising:
determining the similarity between the user portrait of the user to be promoted and the user portrait of each user according to the user tags in the user portrait and the weights corresponding to the user tags;
and pushing the promotion content to the user with the similarity of the user portrait of the user to be promoted larger than a first threshold value.
6. The method of any of claims 1 to 5, further comprising, prior to said determining a user to promote based on the keywords of the promotional content and the user tags in the user representation:
and acquiring the initial popularization crowd category of the popularization content, and pushing the popularization content to users belonging to the initial popularization crowd category.
7. An apparatus for content promotion, comprising:
the system comprises an acquisition unit, a search unit and a search unit, wherein the acquisition unit is used for acquiring popularization content and extracting keywords of the popularization content;
the processing unit is used for determining a user to be promoted according to the keyword of the promotion content and the user label in the user portrait; the user portrait comprises user labels and weights corresponding to the user labels, wherein the weights corresponding to the user labels are determined according to user identification weights, time weights, content weights and behavior weights of websites browsed by users; pushing the promotion content to the user to be promoted;
the processing unit is specifically configured to:
acquiring browsing content, user identification and browsing time of the website browsed by the user;
determining the user tag and a content weight corresponding to the user tag according to the browsing content;
determining the user identification weight corresponding to the user label according to the user identification and the browsing time corresponding to the user identification;
determining a time weight and a behavior weight corresponding to the user tag according to the browsing time corresponding to the user tag;
determining the weight corresponding to the user label according to the user identification weight, the time weight, the content weight and the behavior weight corresponding to the user label;
the weight corresponding to the user label can be shown as formula (9);
the formula (9) is:
wherein the content of the first and second substances,
for user UiUser tag ljCorresponding user tag weights; obji(lj) For user UiUser tag ljCorresponding subscriber identity weight, Ci(lj) For user UiUser tag ljThe corresponding weight of the content is set to,
for user UiUser tag l at kth sub-periodjThe corresponding weight of the time is given to,
for user UiUser tag l at kth sub-periodjA corresponding collection weight;
for user UiUser tag l at kth sub-periodjCorresponding scoring weights;
for user UiUser tag l at kth sub-periodjCorresponding sharing weights; wherein j is 1,2, … …, sumi。
8. A computing device, comprising:
a memory for storing program instructions;
a processor for calling program instructions stored in said memory to execute the method of any one of claims 1 to 6 in accordance with the obtained program.
9. A computer-readable non-transitory storage medium including computer-readable instructions which, when read and executed by a computer, cause the computer to perform the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811081806.2A CN109325179B (en) | 2018-09-17 | 2018-09-17 | Content promotion method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811081806.2A CN109325179B (en) | 2018-09-17 | 2018-09-17 | Content promotion method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109325179A CN109325179A (en) | 2019-02-12 |
| CN109325179B true CN109325179B (en) | 2020-12-04 |
Family
ID=65266047
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811081806.2A Active CN109325179B (en) | 2018-09-17 | 2018-09-17 | Content promotion method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325179B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114861074A (en) * | 2022-07-07 | 2022-08-05 | 深圳市乐易网络股份有限公司 | User data analysis method and system |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110033316A (en) * | 2019-03-22 | 2019-07-19 | 微梦创科网络科技（中国）有限公司 | A kind of target launches the determination method, device and equipment of account |
| CN110222191B (en) * | 2019-04-19 | 2023-08-22 | 平安科技（深圳）有限公司 | User interest portrait construction method, device, computer equipment and computer storage medium |
| CN110310163A (en) * | 2019-07-11 | 2019-10-08 | 易联众民生(厦门)科技有限公司 | A kind of accurate method, equipment and readable medium for formulating marketing strategy |
| CN110880013A (en) * | 2019-08-02 | 2020-03-13 | 华为技术有限公司 | Text recognition method and device |
| CN110659417A (en) * | 2019-09-12 | 2020-01-07 | 广东浪潮大数据研究有限公司 | Information pushing method and system, electronic equipment and storage medium |
| CN111324724B (en) * | 2020-02-13 | 2023-04-11 | 腾讯科技（深圳）有限公司 | Data processing method and device, electronic equipment and computer readable storage medium |
| CN113723984A (en) * | 2021-03-03 | 2021-11-30 | 京东城市(北京)数字科技有限公司 | Method and device for acquiring crowd consumption portrait information and storage medium |
| CN113643070A (en) * | 2021-08-20 | 2021-11-12 | 林秀珍 | Intelligent information pushing method and system based on big data |
| CN115130007B (en) * | 2022-08-29 | 2022-11-15 | 深圳市亲邻科技有限公司 | Brand promotion method and system based on user scene positioning |
| CN116881574B (en) * | 2023-09-07 | 2023-11-28 | 中科数创(北京)数字传媒有限公司 | Directional science popularization pushing method and system based on user portrait |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103024585A (en) * | 2012-12-28 | 2013-04-03 | Tcl集团股份有限公司 | Program recommendation system, program recommendation method and terminal equipment |
| CN103366299A (en) * | 2012-03-30 | 2013-10-23 | 索尼公司 | A method for recommending items and a recommendation system |
| WO2014009400A1 (en) * | 2012-07-13 | 2014-01-16 | Telefonica, S.A. | A method and a system for generating context-based content recommendations to users |
| CN104935970A (en) * | 2015-07-09 | 2015-09-23 | 三星电子（中国）研发中心 | Method for recommending television content and television client |
| CN106815297A (en) * | 2016-12-09 | 2017-06-09 | 宁波大学 | A kind of academic resources recommendation service system and method |
| CN107025310A (en) * | 2017-05-17 | 2017-08-08 | 长春嘉诚信息技术股份有限公司 | A kind of automatic news in real time recommends method |
| CN108133011A (en) * | 2017-12-22 | 2018-06-08 | 新奥（中国）燃气投资有限公司 | A kind of message push method and device |
| CN108153857A (en) * | 2017-12-22 | 2018-06-12 | 北京奇虎科技有限公司 | A kind of method and system for being used to be associated network access data processing |
| CN108536694A (en) * | 2017-03-02 | 2018-09-14 | 广州市动景计算机科技有限公司 | Estimation method, device and the terminal device of user preference |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103870505B (en) * | 2012-12-17 | 2017-10-27 | 阿里巴巴集团控股有限公司 | One kind inquiry words recommending method and query word commending system |
| CN104602042B (en) * | 2014-12-31 | 2017-11-03 | 合一网络技术(北京)有限公司 | Label setting method based on user behavior |
| JP6434188B2 (en) * | 2015-07-27 | 2018-12-05 | マスターカード インターナシヨナル インコーポレーテツド | System and method for tracking data using a data tag provided by a user |
-
2018
- 2018-09-17 CN CN201811081806.2A patent/CN109325179B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103366299A (en) * | 2012-03-30 | 2013-10-23 | 索尼公司 | A method for recommending items and a recommendation system |
| WO2014009400A1 (en) * | 2012-07-13 | 2014-01-16 | Telefonica, S.A. | A method and a system for generating context-based content recommendations to users |
| CN103024585A (en) * | 2012-12-28 | 2013-04-03 | Tcl集团股份有限公司 | Program recommendation system, program recommendation method and terminal equipment |
| CN104935970A (en) * | 2015-07-09 | 2015-09-23 | 三星电子（中国）研发中心 | Method for recommending television content and television client |
| CN106815297A (en) * | 2016-12-09 | 2017-06-09 | 宁波大学 | A kind of academic resources recommendation service system and method |
| CN108536694A (en) * | 2017-03-02 | 2018-09-14 | 广州市动景计算机科技有限公司 | Estimation method, device and the terminal device of user preference |
| CN107025310A (en) * | 2017-05-17 | 2017-08-08 | 长春嘉诚信息技术股份有限公司 | A kind of automatic news in real time recommends method |
| CN108133011A (en) * | 2017-12-22 | 2018-06-08 | 新奥（中国）燃气投资有限公司 | A kind of message push method and device |
| CN108153857A (en) * | 2017-12-22 | 2018-06-12 | 北京奇虎科技有限公司 | A kind of method and system for being used to be associated network access data processing |
Non-Patent Citations (1)
| Title |
|---|
| 基于用户画像的信息智能推送方法;姜建武等;《微型机与应用》;20161230;第35卷(第23期);86-89、92 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114861074A (en) * | 2022-07-07 | 2022-08-05 | 深圳市乐易网络股份有限公司 | User data analysis method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109325179A (en) | 2019-02-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325179B (en) | Content promotion method and device | |
| US20190197416A1 (en) | Information recommendation method, apparatus, and server based on user data in an online forum | |
| JP7356206B2 (en) | Content recommendation and display | |
| CN105989004B (en) | Information delivery preprocessing method and device | |
| JP6301958B2 (en) | Method and apparatus for configuring search terms, delivering advertisements, and retrieving product information | |
| TWI512508B (en) | Recommended methods and systems for recommending information | |
| US20190332602A1 (en) | Method of data query based on evaluation and device | |
| CN108537578A (en) | Advertisement sending method based on big data and device | |
| CN106445963B (en) | Advertisement index keyword automatic generation method and device of APP platform | |
| CN108805598B (en) | Similarity information determination method, server and computer-readable storage medium | |
| EP2941724A1 (en) | Method and apparatus for generating webpage content | |
| CN113744016B (en) | Object recommendation method and device, equipment and storage medium | |
| CN108256537A (en) | A kind of user gender prediction method and system | |
| CN102929939A (en) | Personalized information supply method and device | |
| CN110929138A (en) | Recommendation information generation method, device, equipment and storage medium | |
| CN107590267A (en) | Information-pushing method and device, terminal and readable storage medium storing program for executing based on picture | |
| CN108229991B (en) | Method and device for displaying aggregation promotion information, browser and terminal equipment | |
| CN110597987A (en) | Search recommendation method and device | |
| US20230231930A1 (en) | Content processing method and apparatus, computer device, and storage medium | |
| KR101441164B1 (en) | Object customization and management system | |
| CN114881685A (en) | Advertisement delivery method, device, electronic device and storage medium | |
| CN110674388A (en) | Mapping method and device for push item, storage medium and terminal equipment | |
| CN112561636A (en) | Recommendation method, recommendation device, terminal equipment and medium | |
| CN107085573B (en) | Hotspot information acquisition method and device | |
| CN116823410A (en) | Data processing method, object processing method, recommending method and computing device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |