JP2014006930A - Adheat advertisement model for social network - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide methods and systems for placing advertisements on a social network.SOLUTION: In an embodiment, a computer-implemented method includes a step of receiving, at a server, information indicating activity levels of users of a computer-implemented social network or acquaintance relationships of the users on the computer-implemented social network. The method further includes a step of generating, by the server, influence scores for the users, based on the received information. The method also includes a step of recursively propagating by the server an advertisement through the computer-implemented social network between users having an acquaintance relationship by transmitting the advertisement from a propagating user to a recipient user when a difference between a first influence score of the propagating user and a second influence score of the recipient user is greater than a threshold.

Description

  This document generally describes a method and system for placing advertisements on social networks.

  The present disclosure relates generally to providing advertisements to users of social networks. A social network may be an online system that provides a forum for interaction between geographically separated users. The aim of social networks can be various aspects of social interaction, such as friendship and business networking.

  Some methods of advertising to social network users target social network and social network users for specific content. In some methods, the entire social network is targeted for advertisements directed to the general purpose (eg, business) of the social network. In other methods, specific users are targeted for advertisements based on the content of each user's profile page. For example, if a user is talking about sports on their profile page, sports-related advertisements may target the user.

JP 2008-305258 A

  This document describes a technique for providing advertisements to users of social networks. In general, viewer contact with a particular advertisement, such as an advertisement that a user skips, can be monitored and the system then has a high degree of similarity to the advertisement that the user has seen so far, and the user has skipped so far. It is possible to select another advertisement that is determined to have a low similarity to the advertisement.

  In one embodiment, a computer-implemented method includes information indicating a user's activity level on a computer-implemented social network or an acquaintance relationship on the user's computer-implemented social network. Including receiving at the server. The method further includes generating a user influence score by the server based on the received information. The method also includes receiving an advertisement from the propagating user and sending it to the user when the difference between the propagating user's first influence score and the receiving user's second influence score exceeds a threshold. To recursively propagate advertisements by a server between users having acquaintance relationships throughout a computer-implemented social network.

  In another embodiment, a system for propagating advertisements across users of a social network includes a server that receives information indicating one or more servers and a user's level of activity on the social network or an acquaintance relationship on the user's social network. Interface to one or more. The system is implemented in one or more servers and means for generating a user's influence score based on the received information, the propagation user's first influence score and the receiving user's second influence score. When the difference between the influence score exceeds a threshold value, the advertisement is recursively propagated across the social network between the users with acquaintances by receiving the advertisement from the propagating user and sending it to the user And a recursive propagation module.

  In another embodiment, a computer-implemented method includes outputting an advertisement to a first user of a computer-implemented social network, the output causing the advertisement to be displayed to the first user. The method includes receiving input from a first user associated with an advertisement, a first user's influence score, and a second user of a social network associated with the first user by an acquaintance relationship. Searching for influence scores. The method further includes propagating the advertisement to a second user having an influence score that is less than the influence score of the first user.

  The details of one or more embodiments are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

1 is an example system that receives bids for advertising opportunities to users of a social network and propagates advertisements from user to user across a social network. FIG. FIG. 4 is another diagram of an example system that receives bids for advertising opportunities to users of a social network and propagates advertisements from user to user across the social network. 6 is a flowchart of an exemplary process that provides an opportunity for a social network user to display an advertisement based on an advertiser request. 6 is a flowchart that individually illustrates exemplary processing for recursively propagating advertisements to all users of a social network. It is a flowchart which shows individually the exemplary process which produces | generates the model of a user's interest from the content which the user produced | generated. 6 is a flowchart that individually illustrates an exemplary process in which an advertiser bids on an opportunity to display an advertisement to a user of a social network. It is a flowchart which shows individually the exemplary process which propagates an advertisement from a propagation user to a receiving user. 1 is a schematic diagram of a computer system.

  Like reference symbols in the various drawings indicate like elements.

  This document generally describes placing advertisements on social networks. Specifically, this document describes the allocation of opportunities to display advertisements to users and the propagation of advertisements across social networks. In some cases, methods and systems are described that allow advertisers to target users of social networks according to their interests and social network influences. Opportunities to display advertisements to interested and influential users related to advertisements are an opportunity for advertisers to have irrelevant interests and non-influential users (or related interests and influences) It may be more valuable than the opportunity to display advertisements to users who do not. A user's social network impact may be determined by noting the level of activity and / or acquaintances of the user's social network. Advertisers can receive a list of anonymous users ranked according to their interests and influences. The bidding mechanism can be used to compromise between multiple advertisers seeking to have the opportunity to display advertisements to a finite number of relevant and influential users on social networks.

  In some cases, once an advertiser is given the opportunity to display an advertisement to a specific anonymous user, the advertisement from the advertiser is propagated from the user to the user's friend using a heat diffusion model. Can be done. For example, a user's influence on a social network can be expressed as heat intensity or heat score, where the greater the user's influence, the higher the heat score. The propagation between users can then be modeled using a thermal diffusion model. For example, as long as the targeted user of an ad has a greater “heat” than a user who has not yet been targeted, the ad spreads between two connected users (propagation) can do. This can result in advertisements propagating through the entire social network from users with high influence to users with low influence. One advantage gained from the above method is that advertisers can maximize advertising efficiency by propagating advertisements from influential users to affected users.

  FIG. 1 is an illustration of an example system 100 that receives bids for an opportunity to advertise to a user of a social network and propagates the advertisement from user to user across a social network. The system 100 may include advertisers 102a-102c, a server 112 that receives bids, and a social network 118.

  Advertiser 102a may submit a request to display advertisement 108 to a user of a social network having certain characteristics 106. Server 112 may receive the request and provide advertiser 102a with a list 114 of anonymous users ranked according to user impact score and / or user similarity to the requested feature. The advertiser 102a can then bid on the opportunity to display the advertisement 108 to one or more of the 114 ranked users. Server 112 may receive bids 104b and 104c from multiple advertisers 102b and 102c for the opportunity to display advertisements to 114 anonymous users. For example, in the bid table 114, there are two bids ($ 1.50 for sports ads and $ 0.60 for music ads) for the opportunity to display advertisements to user U4. Server 112 can then give the highest bidder an opportunity to display the advertisement, at which point advertisement 116 may be displayed on social network 118.

  The advertisement may first be displayed to an anonymously bid user (eg, user U4 120). The advertisement 116 may then be propagated throughout the social network 118 along with connections between users. For example, as long as the propagating user's thermal score or influence score exceeds the receiving user's thermal score, the advertisement may propagate from a user associated with another user. For example, user U4 120 is tied to user U17 122. At 114, user U4 has a heat score of 500 and user U17 has a heat score of 450. The advertisement can propagate from user U4 120 to user U17 122 because U4's thermal score exceeds U17. However, at 114, the heat score for U2 is 510, which exceeds the heat score for user U4, so the advertisement cannot propagate from user U4 120 to user U2 126.

  FIG. 2 is another diagram of an example system 200 that receives bids for advertising opportunities to users of social networks and propagates advertisements from user to user across social networks. The system 200 can include a client side 202 and a server side 204. The client side 202 can interface with the advertiser 206 to submit a request and bid to display the advertisement to a selected user of the social network. In addition, the client side 202 can display an advertisement submitted by the advertiser 206 as an interface of the user 208, for example, in a page viewed by the user on the social network. Server side 204 can receive the request of advertiser 206 and provide the selected user 208 the opportunity to display the advertisement through the bidding process. Server-side 204 also uses a heat diffusion model between social network users 208 that are connected by, for example, acquaintances (e.g., users are designated as friends, users visit each other's profiles, etc.) The advertisement can be propagated.

  The client side 202 can include an advertiser 206, which submits the characteristics of the user who wants to target the advertisement. An advertising interface, referred to herein as the Adheat interface 210, can receive the features and send them to the server side 204. The server side 204 can include a web server 214 that receives features and uses the database 216 to communicate each user's influence on the social network and the features submitted by the advertiser. Create a list of users ranked by similarity. After ranking, in some embodiments, the web server 214 can send the ranked list to the client side 202.

  In some embodiments, the web server 214 can create an ordered list of users in response to receiving advertisements and / or advertisement content from the advertiser 206. In an ordered list, users can be ordered according to the likelihood that the user will accept the content of the advertisement and / or the amount of influence the user has on other users of the social network. Users listed near the top of the ordered list are more likely to accept advertising content than users listed near the bottom of the ordered list and / or other users of the social network Has a great influence on

  The client-side Adheat interface 210 can receive a ranked list of users and provide it to the advertiser 206. The advertiser 206 can provide the Adheat interface 210 with a bid for the opportunity to display the advertisement to a selected user on the ranked list. The Adheat interface 210 can then send the bid to the server side 204. After receiving the bid, the web server 214 may auction and give the ranked user the opportunity to display the advertisement based on the bid. The auction process is subsequently described in more detail.

  The web server 214 can send the advertisement to the client side 202. The gadget 212 on the client side 202 can display to the user 208 an advertisement provided by the advertiser 206 that has won the opportunity to display to the user 208. The advertisement can then be propagated throughout the social network, represented here using the friend graph 218. This is explained in more detail below.

  As described above, the advertiser 206 can provide the Adheat module 210 with a request to target users of social networks for advertising. The request may include information specifying the content of the advertisement and / or the user characteristics that the advertiser 206 desires to target the advertisement. User characteristics may include personnel information such as the user's age or age range, gender, interest, occupation, income, educational background, or geographic location.

  In some cases, advertiser 206 may provide Adheat module 210 with one or more keywords related to the user that the advertiser wants to target. For example, the advertiser 206 can provide the Adheat module 210 with a request to display an advertisement related to “sports” to “male” users “aged between 18 and 35”.

  In some embodiments, the advertiser 206 can select from the drop-down menu descriptors for the content of the advertisement and for the user characteristics that the advertiser 206 wants to target the advertisement. For example, the drop-down box can give the advertiser multiple user characteristics for selection. In some cases, an advertiser may provide the advertisement itself to the Adheat module 210, which can analyze the advertisement and determine the content of the advertisement. The advertisements provided to the Adheat module 210 may be text, images, videos, or any combination thereof.

  The Adheat module 210 can determine the content of the advertisement using a content modeling method that derives the content of the advertisement based on words and / or symbols used and / or associated with the advertisement. In some cases, the content of the advertisement may be derived from the text of the advertisement. If advertisements are provided for text or video, optical character recognition can be used to create advertising text. In some cases, the content of the advertisement may be derived from meta information associated with the advertisement.

  Adheat module 210 can submit a request from advertiser 206 to web server 214. In response, the web server 214 can provide the Adheat module 210 with a list of ranked anonymous users that match one or more of the characteristics specified by the advertiser 206.

  The opportunity to display advertisements to users ranked at the top of the list may be more valuable to advertiser 206 than to display to users ranked at the bottom of the list. Users ranked at the top of the list may be more influential on social networks and may be closer to the features specified by advertiser 206. The ranked list of users includes information about each user, such as anonymized user identifiers (which prevent advertisers from identifying specific users on social networks), influence scores, and relevance scores. obtain. Users can be ranked on the list according to influence score, relevance score, or a combination thereof. In general, the higher the user's influence score and relevance score, the more likely the user will be presented at the top of the ranked list.

  The influence score may be a measure of the user's influence on other users of the social network. The impact score may be based on the user's activity level and / or acquaintance. The user activity level may be the degree of user interaction on the social network.

  An acquaintance relationship can be a connection between two users on a social network, such as a friendship. A user's influence score may be influenced by the amount and / or importance of user acquaintances.

  For example, a user's level of activity may include the quality and / or amount of user interaction with other users on the social network, the quality and / or amount of content generated by users on the social network, and / or within a period of time. Can be determined by the frequency of user interaction and / or content generation on social networks. A user with a high quality and / or amount of interacting and / or generated content may be determined to have a higher activity level than a user with a low quality and / or amount of interacting and / or generated content. Also, users with a high frequency of exchange and / or generated content during a period can be determined to have a higher activity level.

  User interaction on social networks means at least posting a comment on another user's page, sending a message to another user, participating in a forum on a social network, adding another user as a friend Posting a comment in a chat room on a social network, and / or browsing another user's profile page. Content generated by the user may include at least uploading an image or video, updating a user profile page, posting an entry on a blog, and / or updating user status information.

  The quality of user interaction on social networks may be determined by evaluating the nature of the interaction and / or the responsiveness of the interaction to the context in which the interaction occurs. For example, the following two comments about sports can be posted. 1) “I like sports” and 2) “I love watching sports, especially watching football and hockey. Photos from the Super Bowl party posted in my profile Please check out. " Since the latter is substantial, it can be determined that the quality is higher.

  Further in the previous example, if one of these two comments appears in response to a topic about the music genre, it is not responding to the context in which they appear (for example, the music genre). They are not likely to be of high quality. Comments that are less substantial but are responsive to the context in which they appear may be determined to be of higher quality than comments that are substantial but off topic. For example, a comment posted on a topic about a music genre that begins with “Blues is my favorite category of music” may be determined to be of higher quality than the second, more substantive comment above.

  Like the quality of user interaction, the quality of user-generated content can also be evaluated for substance and responsiveness to the context. Also, the quality of content generated by the user can be evaluated based on originality. Original user-generated content (eg, previously not posted on a social network by another user) is better than non-original content (eg, copied from another user's profile page) Can be determined.

  The amount of user interaction on the social network and the content generated by the user can be determined by the amount of content posted on the plurality of interaction and social networks. For example, it may be determined that the user posting five comments has more activity than the user posting two comments.

  In some cases, a user's activity level may be determined by a combination of user interaction and the quality and quantity of content generated by the user. The activity level can be determined by a weighted amount of user interaction and / or a weighted amount of content generated by the user, where each interaction and / or user-generated content is weighted according to its quality. For example, when the user posts two comments similar to comment 2 (above), the activity level may be higher than when the user posts five comments similar to comment 1 (above).

  Also, a user's activity level can be based at least in part on the frequency of interaction within a period and / or content generated by the user. For example, a user who interacted with 5 other users last month and posted 10 new photos (e.g. content generation) interacted with only 2 other users last month and only 5 new photos It can be determined that the activity level is higher than that of another user who has posted only one sheet. The exchanged and / or generated content can also be weighted according to the determined quality.

  A user's influence score may be influenced by the number of acquaintances of the user. Users with many friends can be more influential than users with few friends. Therefore, a user who has many acquaintance relationships may have a greater influence score than a user who has few acquaintance relationships.

  The user's influence score can be influenced by the importance of the user's acquaintance relationship. Users with more important acquaintances may have greater social network influence than users with less important acquaintances. The importance of the user's acquaintance is based on the influence score of the user's friend (the user who is connected to the user by the acquaintance) and / or whether the acquaintance is tied to a separate subgraph of the social network Can be based on.

  In some cases, a user's acquaintance relationship with a user with a large influence score may be more important than an acquaintance relationship with a user with a small influence score. In other cases, a user's acquaintance relationship with a user with a small influence score may be more important than an acquaintance relationship with a user with a large influence score.

  The user's social network may be represented as a graph of users (nodes) connected to each other by acquaintance relationships (edges), as illustrated in the friendship graph 218. An acquaintance relationship can be an explicit relationship between users of a social network. For example, a friend designation agreed between both users between two users may be obvious. Acquaintances can be implicit. For example, even though there is no obvious friendship between users U3 and U4, user U3 visits the user U4 profile page, and user U3 is via an interface on a social network (e.g. chat room, email exchange, etc.) If you interact with user U4 and / or if user U3 and user U4 both have a common friend (for example, they both have a clear friendship with user U5), then they know each other between user U3 and U4 Can exist implicitly.

  In the social network graph, the sub-graph may be a group of users, and the user group has many acquaintance relationships with each other, but has few acquaintance relationships with users other than the group. The acquaintance relationship that connects users in two separate subgraphs may be more important than the acquaintance relationship that connects users in the same subgraph. For example, users U2, U4, U6, and U17 of social network 118 are linked through an acquaintance relationship in the same subgraph. The connection between user U10 and user U4 that are part of a subgraph that is separate from the subgraph of U4 may be considered as connecting users of separate subgraphs.

  The relevance score can be a measure of the similarity between a user feature and the user feature that the advertiser 206 has requested to be targeted by the advertisement. As already described with respect to advertiser 206 requirements, the user characteristics may include personnel information such as the user's age or age range, gender, interest, occupation, income, educational background, or geographic location. Users with features similar to those requested by advertiser 206 may have a higher relevance score than users with less similar features. For example, suppose user U3 is “male, 25 years old, likes hockey” and user U5 is “male, 37 years old, likes blues music”. In the case of an advertiser request to display an advertisement to a “male” user who is “18-35 years old” who likes “sports”, the user U3 has a higher relevance score than the user U5 for that request.

  In some cases, the user activity level score may be determined in view of the relevance of each interaction to the advertiser 206 request. In such an example, each interaction may be weighted according to its relevance to the feature requested by advertiser 206. For example, using the scenario from the previous stage, users U3 and U5 post the same number of substantive comments on the social network, but if many of user U5's posts are related to sports, user U5 Also, the weight of activity level score is increased.

  In some embodiments, the web server 214 can receive a ranked list of users by querying a database 216 hosted on the server side 204. Database 216 may include anonymized information about users of social networks. For example, anonymized information may include information regarding user characteristics, social network interactions, social network influence, and / or relevance to advertiser 206 requests.

  In some cases, the database 216 may store models for users of social networks. The model can provide a summary of user features, including interests, in a maintained and queried form. The model can generate information that is inserted into the database 216 by the web server 214 or another component not shown on the server side 204. For example, latent topic modeling can be used to generate a model. Latent topic modeling can create models by processing data from social networks (eg, user profile pages, user interactions) to derive user characteristics (eg, interests).

  Upon receiving a ranked list of users from the web server 214, the Adheat module 210 can provide the advertiser 206 with a ranked list of users. Advertiser 206 can then provide Adheat module 210 with a bid for the opportunity to display the advertisement to one or more users included in the ranked list of users. Bidding may specify a minimum bid per user, a maximum bid per user, and / or a total budget for propagating advertisements across the social network. Advertiser 206 can bid on individual ranked users and / or groups of users in a ranked list.

  The opportunity to display advertisements to users that are highly ranked on the ranked list can be more expensive than the opportunity to display advertisements to users that are ranked low. This can happen when users ranked at the top of the list are highly influential on social networks and relate to features submitted for advertisements by advertiser 206. Advertisers 206 who want to win the opportunity to display ads to users who are ranked at the top of the list are likely to have to pay higher bids for those users than those who are ranked low This is because other advertisers are more likely to bid on the same influential user.

  In some embodiments, the Adheat module 210 may perform some or all of the functions performed by the web server 214 with respect to ranking social network users. For example, in response to receiving an advertisement request from advertiser 206, Adheat module 210 can send a request for a list of users to web server 214 without sending information about the content of the advertisement of advertiser 206. . In response, the web server 214 can provide the Adheat module 210 with a list of users and associated information (eg, user impact score, user interest). The Adheat module 210 can then compute the user's ranking based on the received user's influence score and / or the similarity between the received user's interest and the content of the advertisement. After computing the ranked list, Adheat module 210 can display the ranked list to advertiser 206 and receive user bids from advertiser 206.

  In such embodiments, the Adheat module 210 can request the web server 214 for users who are in a certain influence score range or some percentile of influence scores. For example, the Adheat module 210 may request the web server 214 to return a user with an impact score greater than 450. In another example, the Adheat module 210 may request the web server 214 to return users with an impact score that is within the top 10% of the impact score. The Adheat module 210 can then rank these users as described above.

  Adheat module 210 can submit bids from advertiser 206 to web server 214. The web server 214 can receive bids for the opportunity to display advertisements to multiple social network users from multiple advertisers 206 through multiple Adheat modules 210 (see, eg, 104a-104c in FIG. 1). The web server 214 can auction for opportunities to display advertisements to users of social networks based on bids received from advertisers 206. An auction can give the highest bidder the opportunity to display advertisements to users of social networks.

  For example, advertiser A1 could submit a maximum bid of $ 1.50 on an opportunity to display an advertisement about sports to anonymous user U1. Advertiser A2 could also submit a maximum bid of $ 1.00 at the same opportunity to display the advertisement to user U1, but advertiser A2's advertisement may be related to music. The auction performed by the web server 214 gives the advertiser A1 an opportunity to display an advertisement to the user U1, because its highest bid is larger.

  After the auction gives the user the opportunity to display the advertisement, the web server 214 can propagate the advertisement to the gadget 212 associated with the user 208. In some cases, gadget 212 is an interface that displays advertisements on pages viewed by users on social networks. Advertisements of advertisers 206 that have won the auction can be sent to gadget 212 (which can display advertisements to user 208 through a computer display). Gadget 212 may provide the user with an interface that allows the user to contact the advertisement. The gadget 212 can be a simple HTML or Java Script application or other application that can be embedded in a web page. In some cases, the user's contact with the advertisement may include the user selecting whether to propagate the advertisement to other users who are associated with the user through acquaintance relationships.

  In some cases, users who come into contact with the advertisement can participate in distributing the profits of the advertisement. In some embodiments, the profit distributed to the user is a percentage of the profit attributable to user interaction, a fixed amount, an amount based on the number of users viewing and / or contacting advertisements resulting from user interaction, And / or an amount based on the characteristics and / or influence of a user viewing and / or contacting an advertisement resulting from user interaction. For example, a user who contacts an advertisement can receive an advertising profit dividend based on the number of users the advertisement is propagated by deciding to propagate the advertisement to other users of the social network.

  In some embodiments, advertisement propagation is attributed to the user only if the user propagates the advertisement directly through one of their acquaintances. In other embodiments, advertisement propagation is attributed to the user when it is possible to trace back to user interaction in a series of advertisement propagation from user to user. In such embodiments, the benefits attributable to the propagating user may decrease as the user's propagation increases. For example, user U1 can receive 10% of the profit by propagating the advertisement to user U2. However, when user U2 propagates the advertisement to user U3, user U1 can receive only 5% of the profit. Similarly, when user U3 propagates the advertisement to user U4, user U1 can receive only 2.5% of the profit.

  The gadget 212 can propagate the advertisement to other users of the social network based on the heat diffusion model. Thermal diffusion is a concept from thermodynamic physics. Generally speaking, thermal diffusion stipulates that when there is a difference in heat between two adjacent particles, heat spreads (diffuses) from one particle to the other. For example, if particles A and B are next to each other and have temperatures of 70 degrees and 50 degrees, respectively, heat is transferred from particles A to particles B. However, heat does not transfer from particle B to particle A. If particles A and B both have a temperature of 70 degrees, no heat transfer will occur.

  By assigning each user a “temperature”, the advertisement is propagated throughout the social network using a thermal diffusion model. Advertisements are sent to the receiving user from the propagating user as long as the propagating user has a higher "temperature" than the receiving user. In some embodiments, the thermal diffusion model uses a thermal index score to represent the user's temperature. The user's thermal index score can be based on the user's social network influence score (above). The user's thermal index score and influence score may have a direct relationship (for example, the thermal index score increases as the user has influence).

  Under the thermal diffusion model structure, advertisements can propagate between “adjacent” users. Two users can be considered next to each other as represented by the connection between the nodes on the friend graph 218 if they are connected to each other on the social network. Connections between users can take the form of acquaintance relationships. The acquaintance relationship may be a relationship that means, among other things, the user is a friend, a colleague, a classmate, and / or a member of the same group. In other embodiments, the acquaintance relationship has users viewed each other's profiles, posted content to the user's profile page, and / or received electronic communications (eg, a chat request from another user). You can show that.

  In some embodiments, the advertisement propagates from the propagating user to all users who are associated with the propagating user and have a lower thermal index score than the propagating user. For example, user U3 is a propagation user with a thermal index score of 70. User U3 is associated with users U5, U8, and U9, and their thermal index scores are 50, 35, and 80, respectively. The advertisement displayed to propagating user U3 spreads to users U5 and U8 because their thermal index score is lower. The advertisement does not propagate to the user U9 because the thermal index score of the user U9 is greater than the propagation user U3. However, when an advertisement is displayed to user U9 and user U9 is functioning as a propagation user, the advertisement displayed to user U9 is propagated to user U3, which is the user's U3 thermal index score. This is because it is lower.

  Under the thermal diffusion model, recursion is used to propagate advertisements across social network users. The advertisement spreads from the propagating user to the receiving user, and each receiving user is associated with the propagating user and has a lower thermal index score than the propagating user. After the advertisement is displayed, each receiving user can function as a propagation user. Thus, the advertisement spreads from the receiving user to other users who are tied to the receiving user and have a lower thermal index score than the receiving user.

  For example, using the above scenario, assume that user U9 starts as a propagation user. The advertisement is displayed to user U9 and then propagated to its associated user with a low thermal index score. In this case, since the heat index score of the user U3 is lower than that of the user U9, the advertisement is transmitted to the user U3. User U3 then acts as a propagation user. With user U3 as the propagating user, the advertisement is then propagated to users U5 and U8 because users U5 and U8 have a lower thermal index score than user U3. Users U5 and U8 then act as propagation users. The advertisement is propagated to a user who has a lower thermal index score than the user U5 and is associated with the user U5. The advertisement is also propagated to a user associated with user U8 who has a lower thermal index score than user U8.

  The recursive propagation of advertisements can continue until termination conditions are met. In some cases, the termination condition may be that the advertising budget is exhausted. In another example, the termination condition may be that the propagating user chooses not to contact the advertisement. In such an example, the gadget 212 can replace the advertisement with an alternative advertisement for display to the user 208. In another example, the termination condition may be that the receiving user's thermal index score exceeds the propagating user's thermal index score. In another example, the termination condition may be that the receiving user has already received an advertisement.

  The gadget 212 may send data regarding contact with the advertisement of the user 208 to the web server 214. The transmitted data may include information regarding advertisement viewing and clicks.

  Web server 214 may consist of one or more servers. Web server 214 may include an interface 220, a request scoring module 222, an advertisement assignment module 224, and a recursive propagation module 226. Interface 220 may allow communication with client side 202. Request scoring module 222 can score bids received from advertiser 206 for the opportunity to display advertisements to at least one social network user. The advertisement assignment module 224 can assign social network users an opportunity to display advertisements based on bids scored by the request scoring module 222. The recursive propagation module 226 can recursively propagate advertisements to users of social networks based on the user influence score.

  FIG. 3 is a flowchart of an example process 300 that provides an opportunity for a social network user to display an advertisement based on an advertiser request. Process 300 may be performed by a system such as, for example, systems 100 and 200, and for the sake of clarity, the following description uses systems 100 and 200 as the basis for an example describing the process. However, another system or combination of systems may be used to perform the process 300.

  Process 300 may begin at step 302 by receiving an advertiser request to target a social network user with a particular feature. Advertiser requests may be received on the server side from the client side. For example, the web server 214 can receive a request from an advertiser 206 on the client side 202.

  In step 304, the process 300 may rank the users of the social network based on the similarity between the user feature and the particular feature and the user impact score. User ranking may be performed in response to a request received from an advertiser. For example, the web server 214 may sort users based on profile information and information submitted by advertiser 206 that specifies the user characteristics that the advertiser wants to target. In some embodiments, the web server 214 can pre-sort or rank users based on influence. For example, the user's influence ranking may be updated every certain period (for example, every hour). Then, when users are ranked according to the features submitted by the advertiser, the pre-calculated influence rank is combined with the feature ranking to create a user ranking based on the user influence and features. Also good.

  In some embodiments, the process 300 can also include sending a ranked list of users to the advertiser who provided the request. Some embodiments may also include receiving advertiser bids for opportunities to display advertisements to users on the ranked list. Advertiser bids may include bids for the opportunity to display advertisements to individual users and / or bids for the opportunity to display advertisements to groups of users. For example, the advertiser can bid $ 1.00 on the occasion of displaying the advertisement to the user U3 and bid $ 2.50 on the occasion of displaying the advertisement to the users U4 to U8.

  In some embodiments, the group of users may be a series of ranked users. For example, an advertiser can bid for an opportunity to display an advertisement to users ranked 1-10 on a ranked list. In some embodiments, the bid amount for a group of users may correspond to an amount for all of the users in the group (eg, $ 1.00 for all users U4-U8). In other embodiments, the bid amount for a group of users may correspond to an amount for each user in the group (eg, $ 1.00 for each user U4-U8).

  The process 300 may score the advertiser's bid for an opportunity to display the advertisement to a user of the social network at step 306. For example, scoring can be based on an advertiser's bid for an opportunity to display an advertisement to a user of a social network. There may be a direct relationship between the resulting score and the bid amount (eg, the larger the bid amount, the greater the resulting score).

  Bids can also be scored based on how well the user has noticed the advertiser's advertisement in the past. For example, the server-side 204 database may store information indicating a history of click-through rates for advertisements submitted by a particular advertiser. Bids from advertisers with high click-through rates may be scored higher than bids from advertisers with low click-through rates.

  In some embodiments, an advertiser's bid for an opportunity to display an advertisement to a group of users may be provided with a single score. In some embodiments, an advertiser's bid for an opportunity to display an advertisement to a group of users may be provided with a score for an individual user of the group. In some embodiments, the score for an individual user of the group may be equal to the other users of the group (eg, score for an individual user = score for the user's group / number of users). In other embodiments, the score for an individual user in the group is an assignment from the score for the group and is commensurate with the user's influence score relative to the other user's influence score. (E.g., score for an individual user = score for a group of users × (individual user influence score / group influence score)).

  In step 308, an opportunity to advertise can be assigned to the advertiser. The allocation of the opportunity to advertise may be based on the advertiser's bid score and / or the score associated with the user on which the advertisement is displayed. Advertisers can be assigned opportunities to advertise to individual users and / or groups of users. Bidding is considered when the allocation of advertising opportunities can be limited to bids received within a particular time frame and / or bids that display advertisements to the user within a particular date range.

  In some embodiments, where an advertiser submits a group bid and the group bid is scored for each user in the group, the opportunity to display the advertisement in the group has a corresponding bid score for each user in the group. Can be assigned to an advertiser if it exceeds bid scores from other advertisers. For example, if an advertiser bids to display an advertisement to users U4-U8 and the bid is scored for each individual user (eg, each user is scored 1.0), other advertisements As long as the primary does not have a bid score greater than 1.0 (below, depending on the scoring technique) for any of the users U4-U8, the advertiser may be given the opportunity to display the advertisement in the group.

  In some other embodiments, where the advertiser submits a group bid and the group bid is scored for each user in the group, the opportunity to display the advertisement to each user in the group is evaluated and individual Can be assigned per user. For example, if an advertiser bids to display an advertisement to users U4-U8 and the bid is scored for each individual user (e.g., each user is scored 1.0), user U4 Opportunities to display advertisements can be assigned to that advertiser unless user U4 has other advertisers with a bid score greater than 1.0. Opportunities for displaying advertisements to users U5-U8 may be similarly assigned regardless of the evaluation and assignment of opportunities to display advertisements to other users in the group.

  In some embodiments, where an advertiser submits a group bid and a group score is provided for the group bid, the opportunity to display in the group may be assigned to the advertiser who has the highest score bid for the group. For example, if a first advertiser and a second advertiser have bid scores of 2.5 and 2.7 on the occasion of displaying an advertisement to users U4 to U8, then the second advertiser will have users U4 to U8. May be given the opportunity to display advertisements.

  In some embodiments, an advertiser submits a group bid and a group score is provided for the group bid. In some embodiments, the minimum score required to display an advertisement to an individual user is determined and the minimum score for the group is aggregated. However, if the aggregated minimum score is less than or equal to the group's bid score, the opportunity can be assigned by assigning the opportunity to the advertiser. For example, the advertiser's bid score for a group of users U4 to U8 is 2.5, and the total minimum score for displaying advertisements in the group is 2.4 (for example, to display advertisements to each of users U4 to U8) Advertisers may be assigned an opportunity to display in a group. The minimum score required to display the advertisement to the user may be determined by checking the maximum bid score for the user and incrementing the maximum bid score by a predetermined amount and / or percentage.

  In some embodiments, there may be more than one opportunity to display the advertisement to the user within a period of time. These opportunities may be assigned to the same advertiser or different advertisers depending on the bid score. For example, if there are three opportunities for an advertisement to be displayed to a user within a certain period of time, three opportunities may be assigned to bids with the three highest scores. For example, four advertisers (A1-A4) bid for the opportunity to display advertisements to four of five users (U1-U5), and three opportunities per user are available Advertisers who bid for the first to third amounts per user are given the opportunity to display advertisements to the user. If A1 bids $ 2.00 to users U1-U4, A2 bids $ 1.50 to users U1-U3, A3 bids $ 1.00 to users U2-U4, and A4 bids $ 0.50 to users U1-U4, the opportunity to advertise Can be assigned to each user as follows. U1-A1, A2, A4; U2-A1, A2, A3; U3-A1, A2, A3; U4-A1, A3, A4; U5-Opportunities are not allocated.

  In some embodiments, when there are multiple opportunities to display an advertisement to a user, if a single advertiser bids more than once among multiple opportunities, the single advertiser Two or more of them can be assigned.

  After assigning an advertiser the opportunity to advertise to one user or multiple users, at block 310, the advertiser's advertisement may be output to the single user or multiple users. The user or users who receive the advertisement may be on the client side, such as a user identified within social network 118 or user 208, for example.

  In some embodiments, users of social networks that have not been assigned an opportunity to advertise, such as user U5 in the previous example, can receive a default advertisement instead of the advertisement assigned to the advertiser. (For example, default advertisements may include advertisements for social networks, profiles of other users on social networks, advertisements for non-profit / charity organizations, etc.).

  FIG. 4 is a flowchart that individually illustrates an exemplary process 400 for recursively propagating advertisements across social network users. Process 400 may be performed by a system such as, for example, systems 100 and 200, and for the sake of clarity, the following description uses systems 100 and 200 as the basis for an example describing the process. However, another system or combination of systems may be used to perform the process 400.

  Process 400 may begin at step 402 and / or step 404. In step 404, process 400 can determine an acquaintance (eg, friendship) relationship between users of the social network. In step 402, the process 400 can determine an activity level (eg, number of posts, friends added, etc.) of a user of the social network.

  At step 406, process 400 may generate a social network user impact score using the determined user acquaintance relationship and the determined user activity level. Combining user acquaintances and activity levels to generate an impact score can vary depending on social network configuration and / or advertiser choice. For example, the first social network may be configured to weight acquaintance relationships twice the activity level, while the second social network may be configured to equally weight acquaintance relationships and activity levels.

  In some embodiments, a social network may have multiple configurations used to generate social network user impact scores. The social network's determination of which of the multiple configurations to use may depend on the state in the social network and / or the type of advertisement displayed. For example, the configuration used by social networks on weekdays may be different from the configuration used on weekends. Different configurations may be determined by the social network based on the personnel information of users accessing the social network at different time periods.

  In some embodiments, the social network may rely on advertiser weighting for acquaintances and activity levels when generating social network user impact scores. The use of influence scores generated by entrusting advertiser weighting can be limited to advertisements from advertisers. Advertisers can provide social networks with weights that apply to specific advertisements from advertisers.

  The generation of social network user impact scores at step 406 may be performed at regular intervals and / or when certain events occur. For example, the impact score is every 10 minutes, every 30 minutes to every hour, every hour, every 6 hours, every 12 hours, every day, every 3 days, every week, and / or every month Can be generated. As another example, an influence score may be generated when a certain number of activity events occur in the social network (e.g., generate an influence score for each of the following events: 1000 on the social network Comments are posted, 10 users enter social networks, 100 acquaintances are formed, etc.).

  In some embodiments, the generation of the social network user's influence score in step 406 may be triggered by displaying an advertisement to the social network user. For example, each time an advertisement is displayed to a user on a social network for the first time, the social network can generate an influence score for the user of the social network. When generation is triggered by an advertisement, the generation of influence scores may be limited to users in the user's subgraph that will be targeted for the advertisement first.

  In some embodiments, the process 400 may display advertisements to users of social networks. For example, referring to FIG. 3 as an overview, an opportunity to display an advertisement to the user may have been given to an advertiser who has made a successful bid through the bidding process.

  At step 408, process 400 may determine whether the propagation user's influence score exceeds the receiving user's influence score. The propagation user may be a user who displays an advertisement. The receiving user may be a user who is associated with the propagation user by an acquaintance relationship (eg, friendship). If the propagating user's influence score exceeds the receiving user's influence score, the process 400 may propagate the advertisement from the propagating user to the receiving user at step 410.

  In some cases, steps 402-406 related to the generation of the influence score may not be performed until step 408. In such cases, steps 402-406 may be performed only for the propagating user and the receiving user. For example, after the advertisement is displayed to the propagating user, the process 400 can determine the influence score of the propagating user and the receiving user.

  In step 410, propagating the advertisement from the propagating user to the receiving user may involve displaying the advertisement to the receiving user. The advertisement displayed to the receiving user may be the same as the advertisement displayed to the propagating user. In step 410, after propagating the advertisement to the receiving user, process 400 may return to step 408. However, when returning to step 408, the receiving user can be a propagation user and a user who is associated with the receiving user by an acquaintance relationship can be a receiving user.

  For example, if user U1 is initially a propagating user, user U2 is initially a receiving user, and U1's influence score exceeds U2's influence score, the advertisement displayed on U1 may propagate to U2 . After the advertisement is propagated to U2, the process 400 can return to step 408, where U2 becomes a propagating user and a user U3 that is associated with U2 by an acquaintance relationship becomes a receiving user.

  Advertisements can be recursively propagated between social network users in steps 408 and 410. Steps 408 and 410 may be performed for each acquaintance relationship of the propagation user. For example, an advertisement displayed to a propagating user may be propagated to each of the propagating user's acquaintance relationships that have an impact score lower than that of the propagating user. Subsequently, each user who receives the advertisement from the propagating user can propagate the advertisement to each of their acquaintance relationships with lower influence scores.

  The recursive propagation of advertisements in steps 408 and 410 may continue until process 400 determines that the propagating user's influence score is less than or equal to the receiving user's influence score. When such a determination is made, the process 400 can end. Recursive propagation can also be terminated when conditions that do not allow continued propagation of advertisements (eg, conditions outlined for FIG. 7) are met.

  In some cases, propagation using a thermal diffusion model can be limited based on the degree of separation between users. For example, if an advertisement is propagated to a user that is more than 6 degrees away from the originating user, the propagation may stop regardless of the adjacent user's thermal index score.

  FIG. 5 is a flowchart individually illustrating an exemplary process 500 for generating a model of user interest from content generated by the user. Process 500 may be performed by a system such as, for example, systems 100 and 200, and for the sake of clarity, the following description uses systems 100 and 200 as the basis for an example describing the process. However, another system or combination of systems may be used to perform the process 500.

  Process 500 may begin at step 502 by collecting information about the user based on the content generated by the user. The collected information may be personnel information (eg, age, gender, geographic location, etc.) and / or may relate to user interests (eg, sports, music, etc.). User generated content includes information that the user provides on his / her own profile page (e.g. interest, age, gender, etc.), the group the user is in (e.g. hiking club group), and / or It may include any content generated by users of social networks, such as comments posted by the user (eg, status updates, comments posted on another user's profile page).

  In some cases, visual and audio media (eg, music, photos, videos) posted by the user may also be collected. Meta tags associated with visual content and audio content may be collected. In some cases, an audio / video content analyzer may be used to derive information about visual content and audio content. For example, the audio / visual analyzer can identify an audio file by comparing a calculated file signature for the audio file to a database of known audio files. In another example, the audio / visual analyzer may attempt to derive information about the content of the photo by looking for items and / or logos in the photo that may be relevant to a particular topic (e.g., in the image Finding a baseball cap, baseball glove, and / or baseball team logo may indicate that the photo is related to baseball).

  In step 504, process 500 can input the collected information to a machine learning algorithm. The machine learning algorithm may be a latent topic modeling algorithm. The latent topic modeling algorithm can examine the collected information and generate a list of topics relevant to the user based on the presented topics. In some embodiments, the algorithm can match the collected information against a database of keywords associated with the topic. These keywords and related topics can be learned and adjusted dynamically by algorithms.

  In step 506, process 500 may generate a topic model that provides the user's interests based on the collected information. A topic model for a user can provide a list of attributes related to the user's interests. The topic model can provide an indication of how interested the user is in a particular topic (eg, the user is most interested in sports and then interested in music). After the topic model is generated, the process 500 can end.

  In generating a topic model, previous topic models may be examined to determine changes in user interest over time. For example, a topic model can provide that the user has recently become interested in the genealogy and that the user's interest in sports has declined over the past year.

  For example, user U10 may have a profile page that includes a blog entry about soccer and skiing and photos from a recent professional soccer game the user went to watch. User U10 may be a member of a video game group on a social network and can post comments regarding the snow conditions of soccer, movies, and ski resorts on other user's profile pages. All of this information is collected by process 500 and input to a machine learning algorithm. Using the machine learning algorithm, process 500 can then generate a topic model for user U10. The topic model can enumerate the following interests, soccer, skiing, movies, and video games in order of greatest interest to least interest. By examining previous topic models for user U10, the topic model is that soccer is consistently the greatest interest of user U10, skiing is a periodic interest that peaks in winter, and movies and Video games can also offer what has been declining over the past year.

  FIG. 6 is a flowchart that individually illustrates an exemplary process 600 for an advertiser to bid on an opportunity to display an advertisement to a user of a social network. Process 600 may be performed by a system such as, for example, systems 100 and 200, and for the sake of clarity, the following description uses systems 100 and 200 as the basis for an example describing the process. However, another system or combination of systems may be used to perform the process 600.

  The process 600 may begin at step 602 by the advertiser entering parameters describing the characteristics of the user that the advertiser wants to target the advertisement. Features can include personnel information (eg, age, gender, geographic location) and / or interests (eg, sports, music). The advertiser can enter this information through an interface on the client side, such as the Adheat interface 210 of the system 200, which can then be submitted to the server side, such as the web server 214 of the system 200.

  At step 604, the process 600 receives a ranked list of users from the server, where the rank is based on the feature similarity and the user impact score. The ranked list of users can be generated on a server, such as web server 214, in a manner similar to that described above with respect to FIGS. The ranked list of users may be received through an interface, such as the Adheat interface 210 of the system 200.

  For example, advertiser A1 may enter the user parameters to be targeted for the advertisement, such as male, age 18-35, interested in sports, and the like. Advertiser A1 can then receive a list of ranked users U1-U5. User U1, ranked at the top of the list, will most closely match the features submitted by advertiser A1 and may also have a high influence score on the social network. User U5 ranked at the bottom of the list may not match the features submitted by advertiser A1 and / or have a low influence score on the social network.

  In step 606, the advertiser can enter the maximum and minimum costs that the advertiser is willing to pay to display the advertisement per user. In some cases, the advertiser can enter the maximum and minimum costs that the advertiser is willing to pay to display the advertisement in a group or group of users. Maximum and minimum costs may be entered through an interface, such as the Adheat interface 210 of the system 200.

  Using the previous example, the maximum and minimum costs that advertiser A1 may be willing to pay for users U1-U5 may be directly related to the rank of each user on the list received at step 604. Advertiser A1 can enter the maximum and minimum costs for user U1 as $ 1.00 and $ 0.50, respectively. However, advertiser A1 may enter maximum and minimum costs of only $ 0.25 and $ 0.05, respectively, for user U5. In some cases, advertiser A1 may not enter bids for lower ranked users, such as user U5.

  In step 608, the advertiser enters an overall budget for displaying the advertisement. The overall budget may be the maximum amount that the advertiser is willing to use for advertisements that are displayed to bidder users and propagated to other users of the social network. Once the budget is exhausted, the advertisement can be stopped from appearing to users of the social network. The overall budget can be entered via an interface, such as the Adheat interface 210 of the system 200.

  In some cases, the budget can be periodic or for a specific period of time. For example, an advertiser may specify that a daily advertisement display budget is $ 50.00. Once the daily budget is exhausted, ads cannot be displayed to users of social networks until the next one-day period begins.

  In step 610, the advertiser can submit his bid (eg, maximum and minimum costs) for the opportunity to display the advertisement to the users provided by the ranked list. Along with bidding, advertisers can also submit an overall budget for advertising. Bids may be submitted by advertisers through an interface, such as the Adheat interface 210 of the system 200, and received by components on the server side, such as the web server 214. After the advertiser submits the bid, the process 600 can end.

  Steps 606-610 can collectively represent a bid process 612 performed by the advertiser for the opportunity to display the advertisement to a user of the social network. Bidding process 612 may be performed periodically by an advertiser for individual advertisements (eg, monthly). In some cases, bidding process 612 may be automated for an advertiser based on a similar advertisement bid previously made by the advertiser. For example, if the advertiser provides average highest and lowest bids to the first and fifth ranked users from $ 1.00 to $ 0.50 and $ 0.25 to $ 0.05, respectively, And these average values of the 5th ranked users can be automatically used as part of a new bid.

  In some cases, automatic bidding may be based at least in part on a user's influence score and a rank in the ranked list of users. For example, automatic bidding may be based on the bid amount per user rank divided by the user impact score (eg, bid amount / (user impact score / user rank)). Since the user influence score and user rank can be provided to the advertiser at step 604, bids can be automatically calculated for each user on the ranked list. In some cases, a calculated bid threshold (e.g., a bid that exceeds $ 0.01), a user influence score (e.g., a score that exceeds the average influence score), and / or user rank (e.g., within the top 25) Rank) may need to be satisfied in order for an automatic bid to be made to the user.

  FIG. 7 is a flowchart individually illustrating an exemplary process 700 for propagating advertisements from a propagating user to a receiving user. Process 700 may be performed by a system such as, for example, systems 100 and 200, and for the sake of clarity, the following description uses systems 100 and 200 as the basis for an example describing the process. However, another system or combination of systems may be used to perform the process 700.

  Process 700 may begin at step 702 by a social network user logging in to the social network. In step 704, the user can view advertisements in the gadget. The gadget may be an interactive interface embedded within the user's view of the social network, such as the gadget 212 described with respect to the system 200. For example, a gadget may be an interactive widget that is displayed within a user's web page-based view of a social network. The gadget can display an advertisement for which the advertiser has made a successful bid for displaying to the user. Gadgets can also display advertisements that are propagated to one user from another user on the social network.

  In step 706, the process 700 can determine whether the user has contacted the advertisement displayed in the gadget. User contact decides to click on the ad, show interest in the ad (e.g., press the “play” button for a video ad), and / or propagate the ad to other users of the social network Can be included. If the user contacts an advertisement displayed in the gadget, process 700 proceeds to step 708. If the user does not contact the advertisement displayed in the gadget, process 700 does not propagate the advertisement in step 718.

  In some embodiments, at step 706, the process 700 approves the user's contact propagation of the advertisement to other users of the social network (e.g., propagation to a user that is associated with the user through an acquaintance relationship). Whether it can be further determined. For example, to cause the advertisement to propagate, the user may need to indicate that they accept the propagation by some sort of contact with the gadget (eg, by clicking on the “Propagate Ad” button). If the user indicates acceptance of the propagation, process 700 can proceed to step 708. Even if the user can contact the advertisement, the process 700 does not propagate the advertisement in step 718 if the user does not accept the advertisement propagation.

  In step 708, process 700 determines whether the advertising budget has expired. As described for process 600, the advertiser can provide an overall budget for displaying the advertisement. Displaying advertisements to users of social networks may prevent the advertisement from being propagated once the budget is exhausted. If the budget has not been exhausted, the process 700 proceeds to step 710. If the budget is exhausted, process 700 does not propagate the advertisement at step 718.

  In step 710, process 700 determines whether the period for displaying the advertisement has expired. The advertiser can specify that the advertisement is displayed within a specific time period. For example, for a music concert advertisement, the advertiser can specify that the advertisement is displayed until the concert time and not displayed thereafter. If the period for displaying the advertisement has not expired, the process 700 proceeds to step 712. If the period for displaying the advertisement ends, the process 700 does not propagate the advertisement in step 718.

  In step 712, process 700 determines whether the advertisement has already been propagated to all users who are associated with the user through acquaintance relationships. As mentioned above, an acquaintance is a connection between two users of a social network, for example, a friendship, posting a comment on another user's profile page, or browsing another user's profile page Interact with another user in an activity promoted by a social network (e.g. playing an online game together through a social network, talking through a chat room or discussion thread), and / or on a social network Belonging to a similar social group or organization. If the advertisement is propagated to at least one of the users associated with the user through an acquaintance relationship, the process 700 proceeds to step 714. If the advertisement has already been propagated to all users connected by the acquaintance relationship, the process 700 does not propagate the advertisement in step 718.

  At step 714, the process 700 determines whether a heat diffusion balance has been reached between the user and a user who has been associated with the user by an acquaintance relationship and the advertisement has not yet propagated. As described above for step 408 of process 400, advertisements may propagate from one user to another as long as the thermal diffusion balance has not been reached. When the propagation user's influence score is less than or equal to the receiving user's influence score, a thermal diffusion balance can be reached. If the thermal diffusion balance has not been reached for at least one user associated with the user through an acquaintance relationship, the process 700 proceeds to step 716. If the heat spread balance has been reached for all users who are associated with the user through acquaintance relationships, the process 700 does not propagate the advertisement in step 718.

  In step 716, the process 700 propagates the advertisement from a user (eg, a propagating user) to a receiving user (eg, a user associated with the propagating user through an acquaintance relationship). The receiving user may be a user who is associated with the propagating user by an acquaintance relationship, has an influence score lower than the propagating user's influence score, and the advertisement has not yet propagated. In some embodiments, the advertisement may be propagated to all receiving users. In other embodiments, advertisements may be propagated only to receiving users that meet the criteria set by the advertiser. For example, the advertiser may indicate that propagation to the receiving user should only occur if the receiving user's influence score exceeds 80% of the propagating user's influence score. After propagating the advertisement to the receiving user, the process 700 can end.

  In step 718, the process 700 does not propagate the advertisement from the user to the user who is connected by the acquaintance relationship. After step 718, process 700 proceeds to step 720. In step 720, the process 700 determines whether there are alternative advertisements available for display in the gadget to the user. The alternative advertisement may be another advertisement that can be displayed to the user (eg, the advertiser has won the opportunity to display the alternative advertisement to the user and the advertisement is propagated to the user). If the replacement advertisement is available for display to the user, the process 700 can return to step 704 and display the replacement advertisement in the gadget to the user. If the alternate advertisement is not available, process 700 can end.

  FIG. 8 is a schematic diagram of a computer system 800. System 800 can be used for the operations described with respect to any of the computer-implemented methods described above, according to one embodiment. System 800 is intended to include various forms of digital computers such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. System 800 may also include mobile devices such as personal digital assistants, cellular phones, smartphones, and other similar computing devices. The system can also include portable storage media such as a universal serial bus (USB) flash drive. For example, a USB flash drive can store an operating system and other applications. A USB flash drive may include input / output components such as a USB connector that may be inserted into a wireless transmitter or a USB port of another computing device.

  The system 800 includes a processor 810, a memory 820, a storage device 830, and an input / output device 840. Each of the components 810, 820, 830, and 840 is connected to each other using a system bus 850. The processor 810 can process instructions for execution in the system 800. The processor may be designed using any of a plurality of architectures. For example, the processor 810 may be a CISC (Compound Instruction Set Computer) processor, a RISC (Reduced Instruction Set Computer) processor, or a MISC (Minimum Instruction Set Computer) processor.

  In one embodiment, processor 810 is a single threaded processor. In another embodiment, processor 810 is a multi-threaded processor. The processor 810 can process instructions stored in the memory 820 or the storage device 830 to display graphical information for a user interface on the input / output device 840.

  Memory 820 stores information within system 800. In one embodiment, memory 820 is a computer readable medium. In one embodiment, the memory 820 is a volatile memory unit. In another embodiment, the memory 820 is a non-volatile memory unit.

  The storage device 830 can implement mass storage for the system 800. In one embodiment, storage device 830 is a computer-readable medium. In a wide variety of embodiments, the storage device 830 can be a floppy disk device, a hard disk device, an optical disk device, or a tape device.

  Input / output device 840 provides input / output operations for system 800. In one embodiment, input / output device 840 includes a keyboard and / or pointing device. In another embodiment, the input / output device 840 includes a display unit for displaying a graphical user interface.

  The functions described may be implemented in digital electronic circuitry or in computer hardware, firmware, software, or combinations thereof. The apparatus may be implemented in a computer program product tangibly embodied in an information carrier, for example, in a machine-readable storage device for execution by a programmable processor, and the method steps may be performed by a programmable processor, It can be implemented by performing the functions of the described embodiments by executing a program of instructions to operate on input data and generate output. The described functions may advantageously be implemented in one or more computer programs, the computer programs being executable on a programmable system including at least one programmable processor, Are coupled to send and receive data and instructions to and from the data storage system, at least one input device, and at least one output device. A computer program is a set of instructions that can be used, directly or indirectly, in a computer to perform an action or produce a result. A computer program may be written in any form of programming language, including a compiler-type language or an interpreted language, and as a stand-alone program or as a module, component, subroutine, or others suitable for use in a computing environment It can be developed into any form including a form as a unit.

  Suitable processors for executing the program of instructions include, by way of example, both general and special purpose microprocessors, as well as one of a single processor or multiple processors of any kind of computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a processor that executes instructions and one or more memories that store instructions and data. Generally, a computer also includes or is operatively coupled to communicate with one or more mass storage devices for storing data files, such devices including internal hard disks and removable disks. , Magneto-optical disks, and magnetic disks such as optical disks. Suitable storage devices for tangibly implementing computer program instructions and data include all forms of non-volatile memory, such as semiconductor memory devices such as EPROM, EEPROM, and flash memory devices, and built-in hard disks And removable disks, magneto-optical disks, and magnetic disks such as CD-ROM and DVD-ROM disks. The processor and memory can be added by or incorporated within an ASIC (application specific integrated circuit).

  To achieve user interaction, the functions include a display device that displays information to the user, such as a CRT (Cathode Ray Tube) or LCD (Liquid Crystal Display) monitor, and a keyboard that allows the user to input to the computer And a computer equipped with a pointing device such as a mouse or a trackball.

  The functionality may be implemented in a computer system, and the computer system comprises a back-end component such as a data server, or comprises a middleware component such as an application server or internet server, or a graphical user interface or internet browser, or A front-end component such as a client computer having any combination of The components of the system can be connected by any form or medium of digital data communication such as a communication network. Examples of communication networks include a local area network (“LAN”), a wide area network (“WAN”), a peer-to-peer network (having ad hoc or fixed members), a grid computing infrastructure, and the Internet.

  The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a network such as those described above. The client-server relationship is caused by computer programs that run on each computer and have a client-server relationship with each other.

  Although several embodiments have been described in detail, other variations are possible. In addition, other mechanisms for providing advertisements to social network users may be used. Also, the logic flows shown in the figures do not require the particular order shown or sequential order to achieve desirable results. Other steps may be provided, or steps may be removed from the described flow, and other descriptions may be added to or removed from the described system. Accordingly, other embodiments are within the scope of the appended claims.

100 system
102a, 102b, 102c Advertiser
104a, 104b, 104c bid
106 Features
108, 116 ads
112 servers
114 bid table
118 social networks
120, 122, 124, 126 users
200 systems
202 Client side
204 Server side
206 Advertiser
208 users
210 Adheat interface, Adheat module
212 Gadget
214 Web server
216 Database
218 Friendship Graph
220 interface
222 Request scoring module
224 Ad allocation module
226 Recursive propagation module
612 bid processing
800 system
810 processor
820 memory
830 storage device
840 input / output device
850 system bus

Claims (23)

A computer-implemented method comprising:
Receiving at a server information indicating an activity level of a user of a computer-implemented social network or an acquaintance relationship of the user on the computer-implemented social network;
Generating an impact score for the user by the server based on the received information;
Determining that the difference between the propagation user's first influence score and the receiving user's second influence score is greater than a threshold;
Sending advertisements from the propagating user to the receiving user causes the server to recursively propagate the advertisements between users having acquaintances across the computer-implemented social network until the termination condition is satisfied by the server Including steps,
Each termination receiving user having an end condition for a specific propagating user and having an acquaintance relationship with the specific propagating user has i) an influence score that is greater than the influence score of the specific propagating user, or ii ) A method comprising: having already received a transmission of the advertisement.   The method of claim 1, wherein the receiving user functions as a second propagating user after the advertisement is transmitted to the receiving user.   The method of claim 1, wherein the termination condition further includes, for a particular propagation user, receiving an indication that the particular propagation user has declined the opportunity to propagate the advertisement throughout the computer-implemented social network. .   The method of claim 3, wherein the indication is generated by a client associated with the particular propagating user in response to an action that is indifferent to the advertisement.   After the user declines to contact, the server replaces the advertisement with a substitute advertisement, and the server recursively substitutes the substitute advertisement between users having acquaintances across the social network. The method of claim 3, further comprising the step of propagating.   The method of claim 1, wherein the termination condition further comprises that the advertising period has expired.   The method of claim 1, wherein the termination condition further comprises that a target number of propagations for the advertisement has been achieved.   The method of claim 1, wherein the termination condition further comprises that the advertising budget has been exhausted.   The method of claim 1, further comprising: providing the propagation user with an opportunity to contact the advertisement by the server.   The method of claim 9, further comprising providing a reward to the propagation user by the server when the propagation user chooses to propagate the advertisement by contacting the advertisement.   The method of claim 10, further comprising calculating a reward amount by the server based on a number of recipient users through which the advertisement is propagated.   The method of claim 9, wherein the opportunity to contact is provided to the propagating user in a gadget located in the propagating user's display of the computer-implemented social network.   The propagation user contacts the advertisement by adding a comment about the advertisement or by explicitly specifying one or more users who will receive the advertisement based on contact with the advertisement. The method described in 1.   The method of claim 1, further comprising obtaining by the server data related to the propagation of the advertisement from the propagating user to each receiving user across the computer-implemented social network.   The method of claim 14, wherein the propagation data includes information related to a number of users viewing the advertisement, a number of users touching the advertisement, or a number of users propagating the advertisement.   Recursively propagating the advertisement throughout the computer-implemented social network determines the threshold by applying a thermal diffusion model to the first influence score and the second influence score. The method of claim 1, further comprising: the thermal diffusion model determining a path by which the advertisement can be propagated throughout the computer-implemented social network. A system that propagates advertisements across social network users,
With one or more servers containing one or more processors,
The one or more processors are:
An interface to one or more of the servers for receiving information indicating a user's activity level on the social network or an acquaintance relationship of the user on the social network;
Means for generating an influence score for the user based on the received information;
Determining that the difference between the first influence score of the propagating user and the second influence score of the receiving user is greater than a threshold, and transmitting an advertisement from the propagating user to the receiving user Through a social network, programmed to provide a recursive propagation module for recursively propagating the advertisement until a termination condition is satisfied between users having an acquaintance relationship,
Each termination receiving user having an end condition for a specific propagating user and having an acquaintance relationship with the specific propagating user has i) an influence score that is greater than the influence score of the specific propagating user, or ii ) A system comprising: receiving a transmission of said advertisement.   18. The system of claim 17, wherein after the advertisement is transmitted to the receiving user, the recursive propagation module allows the receiving user to function as a second propagating user.   The system of claim 17, wherein the propagation module terminates the recursive propagation of the advertisement across the social network. A computer-implemented method comprising:
Outputting an advertisement to a first user of a computer-implemented social network, wherein the advertisement is displayed to the first user by output;
Receiving input from the first user related to the advertisement;
Searching the influence score of the first user, and the influence score of the second user of the social network associated with the first user by an acquaintance relationship;
Determining that a difference between the influence score of the first user and the influence score of the second user is greater than a threshold;
Recursively propagating the advertisement across the computer-implemented social network to the second user until a termination condition is satisfied;
Each termination receiving user having an end condition for a specific propagating user and having an acquaintance relationship with the specific propagating user has i) an influence score that is greater than the influence score of the specific propagating user, or ii ) A method comprising: having already received a transmission of the advertisement.   Further comprising generating the influence score of the first user and the influence score of the second user of the computer-implemented social network associated with the first user by an acquaintance relationship. Item 21. The method according to Item 20.   Generating a user's influence score the user has with respect to other users of the computer-implemented social network based on the user's acquaintance and the user's activity level on the computer-implemented social network The method of claim 21, comprising calculating an influence magnitude.   21. The method of claim 20, wherein the input from the first user includes an indication that the first user wants the advertisement to be propagated to the second user of the social network. .