Semantic query suggestion using Twitter Entities

There are many web information management methods and techniques that help search engines and news services to provide useful suggestions with respect to queries, thus facilitating the users’ search. However, the penetration of microblogging services in our daily life demands to also consider social sphere as far as query suggestion is concerned. Towards this direction, we introduce an algorithmic approach capable of creating a dynamic query suggestion set, which consist of the most viral and trendy Twitter Entities (that is hashtags, user mentions and URLs) with respect to a user׳s query. For evaluation purposes, we firstly compare the results derived from two case studies, against the suggestions of popular services like Google News, Yahoo! News, Bing News, and Reuters. In addition we further evaluate our approach with subjective user ratings against Google Trends service. Finally, we provide comparative results that clearly show that our proposal outperforms other methods and baselines in the respective literature.