DISCOVERING SHARED INTERESTS USING GRAPH ANALYSIS.

The article focuses on the problem of discovering users with particular interests or experience using graph analysis. When users collaborate, they often form networks of colleagues according to shared interests or responsibilities, both within and across bureaucratic boundaries. A specialization subgraph (SSG) of a communication graph is a subset of nodes and edges, where each of the people (nodes) share a common interest, which is reflected in the individual communications (edges). While simply communicating with another person does not imply shared interests, being in an SSG with that person does imply shared interests. In general, a person will belong to many different SSGs according to his or her interests and responsibilities and can try to locate information about a particular topic or accomplish a specific task by consulting an appropriate SSG. The densely interconnected nature of communication graphs allows one to perform shared-interest analysis effectively on data collected at even a modest number of locations.