Exploiting User Signals and Stochastic Models to Improve Information Retrieval Systems and Evaluation

Abstract

Progress and innovation are driven by experiments, but experimentation is useless without an objective evaluation measure that allow researchers to detect the improvements and identify the successful strategies. Due to the experimental nature of Information Retrieval (IR), accurately interpreting the result of a system in terms of user satisfaction is fundamental to push the research in the correct direction. Therefore, measuring systems effectiveness continues to be an active area of research and discussion in the scientific community. It is also the case of this thesis, whose leitmotiv is an investigation of effectiveness measures exploited in different aspects of IR.