Browse Prior Art Database

Method and System for Discovering and Summarizing Communities of Similar Interests in Social Media

IP.com Disclosure Number: IPCOM000237429D
Publication Date: 2014-Jun-18
Document File: 4 page(s) / 491K

Publishing Venue

The IP.com Prior Art Database

Related People

Lyndon Kennedy: INVENTOR [+2]

Abstract

A method and system is disclosed for discovering and summarizing communities of users on social media services which share similar interests. The method and system includes constructing a social network graph between users by linking individuals who have displayed interest on similar pieces of shared media.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Method and System for Discovering and Summarizing Communities of Similar Interests in Social Media

Abstract

A method and system is disclosed for discovering and summarizing communities of users on social media services which share similar interests.  The method and system includes constructing a social network graph between users by linking individuals who have displayed interest on similar pieces of shared media.

Description

Disclosed is a method and system for discovering and summarizing communities of users on social media services which share similar interests.  The method and system includes constructing a social network graph between users by linking individuals who have displayed interest on similar pieces of shared media.

In accordance with the method and system, data related to users who have displayed interest on similar piece of shared media by “liking” or marking the shared media as favorite are obtained.  A graph of such users is constructed wherein the graph displays connection between such users.  Further, the graph is mined and clique percolation method (CPM) is applied on the graph to discover interest communities comprising of users who are connected around a single shared interest as illustrated in Fig. 1.

Figure 1

Initially, collection of users, photographs, and users’ “favorite” behavior around such photographs is collected.  Action of two users displaying inertest for a same piece of media is captured.  This assists in developing an implied social relationship between the two users.  An explicit social network may be extracted from these actions.  Further, CPM is applied to the explicit social network to discover overlapping social communities.  Different colors indicate different communities.  Users with multiple colors belong to multiple communities.  Since communities are formed around similar media, summarization of community interests is displayed.

Additionally, the CPM method is also used to discover overlapping communities in complex networks as illustrated in Fig. 2.  In the CPM method, a k-clique is defined as a set of k fully-connected nodes.  For example, if k=3, then clique is a triangle.  Two k-cliques are said to be “adjacent” if they both contain k – 1 nodes in common.  A community is then formed by merging cliques joined by k – 1 = 2 vertices.  A series of cliques may be reached through any series of adjacent k-cliques.

Figure 2

The overlapping communities may be used for recommending to users to follow or to connect with other users within a community. 

Additionally, the discovered overlapping communities may be analyzed to examine media objects that connect each community.  A network of users and media objects may be displayed as a bipartite graph as illustrated in Fig. 3.

Figure 3

Further, set of users on a per-community basis is sub-divided wherein links from within a community are of different significance for evaluating representativeness of a given photo for a g...