Inferring Social and Psychological Meaning in Social Media

The rapid ascent and adoption of social media as the dominant form of communication has led to significant analysis challenges. In particular, there is a gap in theoretically informed methods to analyze data mined and collected from social media. One primary cause of this gap is the lack of fusing structural sources of information (e.g. social network topology) with non-structural sources (message content, demographics). As a result significant social patterns and meaning that should be derived by goes undetected. Securboration Inc., teaming with Amazon Fellowship recipient Dr. Carlos Guestrin from the University of Washington; and Mr. Thomas Morello from the commercial marketing company Alliance Data Systems, proposes to address this gap. Our solution, referred to as SURF (Social-media Understanding and Reasoning Framework), is a service-based suite of analytical tools that uses structured and non-structured social media information to identify significant topologies, motifs, embedded online communities, and individual features. SURF leverages social media research on reciprocity, degrees of separation, and quantitative comparisons; and extends it based on noise reduction and link analysis research performed by Dr. Guestrin and semantic content analysis techniques pioneered by Securboration. SURF will provide improved algorithms and tools for mining social media that spans domains from marketing to military intelligence.