Towards Personalized Search for Tweets

Abstract

Powerful search capabilities are fundamentally important for micro-blog-based information systems such as Twitter. While recently there has been some works aimed at enhancing the scalability of micro-blog search, very few existing techniques incorporate personalization into their search and ranking processes. This paper argues that since Twitter is a social network (SN)-based micro-blog system, it is essential to personalize search results taking into account the social relationships among various users. In this paper, we outline a scalable and personalized tweet search framework that takes into account the search parameters, the distances of the follower relationships, and the temporal aspects of the tweets when ranking the search results.