Is Document Frequency important for PRF?
Stéphane Clinchant, Eric Gaussier
We introduce in this paper a new heuristic constraint for
PRF models, referred to as the Document Frequency (DF) constraint,
which is validated through a series of experiments with an oracle.We then
analyze, from a theoretical point of view, state-of-the-art PRF models
according to their relation with this constraint. This analysis reveals that
the standard mixture model for PRF in the language modeling family
does not satisfy the DF constraint on the contrary to several recently
proposed models. Lastly, we perform tests, which further validate the
constraint, with a simple family of tf-idf functions based on a parameter controlling the satisfaction of the DF constraint.
ICTIR, 3rd International Conference on the Theory of Information Retrieval, Bertinoro, Italy, 12-14 September 2011.
2011-045ictir_fba.pdf (393.32 kB)