Identifying Time Intervals of Interest to Queries
Dhruv Gupta, Klaus Berberich · 2014
We investigate how time intervals of interest to a query can be identified automatically based on pseudo-relevant documents, taking into account both their publication dates and temporal expressions from their contents. Our approach is based on a generative model and is able to determine time intervals at different temporal granularities (e.g., day, month, or year). We evaluate our approach on twenty years' worth of newspaper articles from The New York Times using two novel testbeds consisting of temporally unambiguous and temporally ambiguous queries, respectively.