In this talk, I will outline various methods of 'mining' the latent temporal information associated with temporally ambiguous or unambiguous queries. The novel contribution lies in detection of interesting periods or points of time associated with queries. This is done by taking into account not only the document metadata but also the temporal information contained in the document body. The methods are formulated, while taking into account a time model that considers the natural uncertainty associated with time. I present a generative model to detect time intervals at different granularity (e.g., day, month, or year) of relevance to the query.