Temporal query classification at different granularities

Abstract

In this work, we consider the problem of classifying time-sensitive queries at different temporal granularities (day, month, and year). Our approach involves performing Bayesian analysis on time intervals of interest obtained from pseudo-relevant documents. Based on the Bayesian analysis we derive several effective features which are used to train a supervised machine learning algorithm for classification. We evaluate our method on a large temporal query workload to show that we can determine the temporal class of a query with high precision.