An Overview on User Profiling in Online Social Networks

Abstract

Advances in Online Social Networks is creating huge data day in and out providing lot of opportunities to its users to express their interest and opinion. Due to the popularity and exposure of social networks, many intruders are using this platform for illegal purposes. Identifying such users is challenging and requires digging huge knowledge out of the data being flown in the social media. This work gives an insight to profile users in online social networks. User Profiles are established based on the behavioral patterns, correlations and activities of the user analyzed from the aggregated data using techniques like clustering, behavioral analysis, content analysis and face detection. Depending on application and purpose, the mechanism used in profiling users varies. Further study on other mechanisms used in profiling users is under the scope of future endeavors.

Cosine similarity-based algorithm for social networking recommendation

“…To profile users based on their behaviors on various social networks, we need the entire accumulated data of all social networking sites, which is hard for researchers to collect. However, based on the activities performed through one particular site for a significant period, the user can be profiled using separate approaches [16].…”