As Time Goes By: Comprehensive Tagging of Textual Phrases with Temporal Scopes

Temporal expressions (TempEx's for short) are increasingly important in search, question answering, information extraction, and more. Techniques for identifying and normalizing explicit temporal expressions work well, but are not designed for and cannot cope with textual phrases that denote named events, such as "Clinton's term as secretary of state". This paper addresses the problem of detecting such temponyms, inferring their temporal scopes, and mapping them to events in a knowledge base if present there. We present methods for this kind of temponym resolution, using an entity- and TempEx-oriented document model and the Yago knowledge base for distant supervision. We develop a family of Integer Linear Programs for jointly inferring temponym mappings to the timeline and knowledge base. This enriches the document representation and also extends the knowledge base by obtaining new alias names for events. Experiments with three different corpora demonstrate the viability of our methods.