Semantic analysis of microposts for efficient people to people interactions

In this paper we present a framework that extracts meaningful knowledge from microposts shared in social platforms in order to build user profiles. This process involves different steps for the analysis of such microposts (extraction of keywords, named entities and their matching to ontological concepts) and their weighting. The concept weighting involves different scores, such as sentiment analysis and statistical patterns which attempt to measure the expertise of the user in the given field. Additionally, we inform on our prototype application, implemented as a social search engine on top of Twitter, which recommends people relevant to a given question.