Probabilistic latent maximal marginal relevance

10 years 6 months ago
Probabilistic latent maximal marginal relevance
Diversity has been heavily motivated in the information retrieval literature as an objective criterion for result sets in search and recommender systems. Perhaps one of the most well-known and most used algorithms for result set diversification is that of Maximal Marginal Relevance (MMR). In this paper, we show that while MMR is somewhat adhoc and motivated from a purely pragmatic perspective, we can derive a more principled variant via probabilistic inference in a latent variable graphical model. This novel derivation presents a formal probabilistic latent view of MMR (PLMMR) that (a) removes the need to manually balance relevance and diversity parameters, (b) shows that specific definitions of relevance and diversity metrics appropriate to MMR emerge naturally, and (c) formally derives variants of latent semantic indexing (LSI) similarity metrics for use in PLMMR. Empirically, PLMMR outperforms MMR with standard term frequency based similarity and diversity metrics since PLMMR ma...
Shengbo Guo, Scott Sanner
Added 18 Aug 2010
Updated 18 Aug 2010
Type Conference
Year 2010
Authors Shengbo Guo, Scott Sanner
Comments (0)