DECAY-BASED RANKING FOR SOCIAL APPLICATION CONTENT
George Papadakis, Claudia Niederée, Wolfgang Nejdl · 2010
Social applications are prone to information explosion, due to the proliferation of user generated content. Locating and retrieving information in their context poses, therefore, a great challenge. Classical information retrieval methods are, however, inadequate in this environment and users inevitably drown in an information flood. In this paper, we present a novel method that facilitates users’ information quests by identifying and improving the accessibility of the most important resources. This is achieved through an information valuation method that estimates how likely it is for each information item to be accessed in the near future. The experiments verify that our method performs significantly better than others typically used in social applications, while being more versatile, too.