Abstract
Time-aware retrieval models exploit one of two time dimensions, namely, (a) <i>publication time</i> or (b) <i>content time</i> (temporal expressions mentioned in documents). We show that the effectiveness for a <i>temporal query</i> (e.g., illinois earthquake 1968) depends significantly on which time dimension is factored into ranking results. Motivated by this, we propose a machine learning approach to select the most suitable time-aware retrieval model for a given temporal query. Our method uses three classes of features obtained from analyzing distributions over two time dimensions, a distribution over terms, and retrieval scores within top-<i>k</i> result documents. Experiments on real-world data with crowdsourced relevance assessments show the potential of our approach.
Users
Please
log in to take part in the discussion (add own reviews or comments).