Leveraging social relevance: Using social networks to enhance literature access and microblog search JURY

An information retrieval system aims at selecting relevant documents that meet user’s information needs expressed with a textual query. For the years 19701980, various theoretical models have been proposed in this direction to represent, on the one hand, documents and queries and on the other hand to match information needs independently of the user. More recently, the arrival of Web 2.0, known also as the social Web, has questioned the effectiveness of these models since they ignore the environment in which the information is located. In fact, the user is no longer a simple consumer of information but also involved in its production. To accelerate the production of information and improve the quality of their work, users tend to exchange documents with their social neighborhood that shares the same interests. It is commonly preferred to obtain information from a direct contact rather than from an anonymous source. Thus, the user, under the influenced of his social environment, gives as much importance to the social prominence of the information as the textual similarity of documents at the query. In order to meet these new prospects, information retrieval is moving towards novel user centric approaches that take into account the social context within the retrieval process. Thus, the new challenge of an information retrieval system is to model the relevance with regards to the social position and the influence of individuals in their community. The second challenge is produce an accurate ranking of relevance that reflects as closely as possible the importance and the social authority of information producers. It is in this specific context that fits our work. Our goal is to estimate the social relevance of documents by integrating the social characteristics of resources as well as relevance metrics as defined in classical information retrieval field. We propose in this work to integrate the social information network in the retrieval process and exploit the social relations between social actors as a source of evidence to measure the relevance of a document in response to a query. Two social information retrieval models have been proposed in different application frameworks: literature access and microblog retrieval. The main contributions of each model are detailed in the following.