An approach for developing intelligent systems for sentiment analysis over social networks

This work presents a novel approach towards building intelligent systems for detecting users' perception on companies brands through the use of social networks. Users' perception is crucial for detecting the possible success or failure of a company in the web 2.0 era more than in the past. For this reason we propose here an approach for building intelligent systems that are able to detect users' opinion about a brand exploring the Twitter social network. By using Twitter messages the proposed approach employs an automatic technique based on semantic differential together with natural language processing and graph ontology exploration for detecting users' perception of brands. We demonstrate how our system could provide companies with knowledge about users' perception for decision making, and knowledge discovery applications, by presenting two use cases.