Fecha: 19 de marzo de 2013

Ponente: Maria-Hendrike Peetz, (Information and Language Processing Systems, UvA)

Lugar de celebración: Sala 1.03, ETSI Informática, UNED (mapa)

Resumen:Temporal information retrieval exploits temporal features of document collections and queries. Temporal document priors are often used to adjust the score of a document based on its publication time. We consider a class of temporal document priors that is inspired by retention functions considered in cognitive psychology; such functions are used to model the decay of memory. We find that many retention functions used as a temporal document prior have a positive effect on overall retrieval performance. We examine the stability of this effect across news and microblog collections and discover interesting differences between retention functions. We also study the problem of optimizing parameters of the retention functions as temporal document priors and again find interesting differences where some retention functions display consistent good performance across large regions of the parameter space. Overall, a retention function based on a Weibull distribution is the preferred choice for a temporal document prior.