Towards a novel and timely search and discovery system using the real-time social web

Abstract

The world of web search is changing. Mainstream search engines like Google and Bing are adding social signals to conventional query-based services while social networks like Twitter and Facebook are adding query-based search to sharing-based services. Our search and discovery system, Yokie, harnesses the wisdom of the crowd of communities of Twitter users to create indexes of proto-content (or recently shared content) that is typically not yet indexed by mainstream search engines. The system includes an architecture [13] for a range of contextual queries and ranking strategies beyond standard relevance. In this paper, we focus on evaluating Yokie's ability to retrieve timely, relevant and exclusive results with which users interacted and found useful, compared to other standard web services. © 2013 Springer-Verlag.