Modeling Expected Utility of Multi-session Information Distillation

An open challenge in information distillation is the evaluation and optimization of the utility of ranked lists with respect to flexible user interactions over multiple sessions. Utility depends on both the relevance and novelty of documents, and the novelty in turn depends on the user interaction history. However, user behavior is non-deterministic. We propose a new probabilistic framework for stochastic modeling of user behavior when browsing multi-session ranked lists, and a novel approximation method for efficient computation of the expected utility over numerous user-interaction patterns. Using this framework, we present the first utility-based evaluation over multi-session search scenarios, using the TDT4 corpus of news stories and compare a state-of-the-art distillation system against a relevance-based retrieval engine. We demonstrate that the distillation system obtains a 44% utility enhancement over the retrieval engine due to multi-session adaptive filtering, accurate novelty detection, and utility-based adjustment of ranked list lengths.