Section: New Results

Modelling and extracting temporal structures

Participants : Pascal Denis, Philippe Muller.

Temporal information has been the focus of recent attention in information extraction. An elegant approach to learning temporal orderings from texts is to formulate this problem as a constraint optimization problem, which can be then given an exact solution using Integer Linear Programming. This works well for cases where the number of possible relations between temporal entities is restricted to the mere precedence relation, but becomes impractical when considering all possible interval relations.

We have proposed this year two innovations [29] , inspired from work on temporal reasoning, that control this combinatorial blow-up, therefore rendering an exact ILP inference viable in the general case. First, we propose to translate the network of constraints from temporal intervals to their end-points, to handle a drastically smaller set of constraints, while preserving the same temporal information. Second, we have show that additional efficiency is gained by enforcing coherence on particular subsets of the entire temporal graphs. We evaluate these innovations through various experiments on TimeBank 1.2 using standard evaluation metrics, and compare our ILP formulations with various baselines and oracle systems.

The evaluation of temporal information extraction, i.e., the comparison of two annotations of a given text, is also a scientific challenge. This is because relations between events in a story are intrinsically interdependent and cannot be evaluated separately. A proper evaluation measure is also crucial in the context of a machine learning approach to the problem. Finding a common comparison referent at the text level is not obvious, and we have argued, in collaboration with Xavier Tannier (LIMSI), in favor of a shift from event based measures to measures on a unique textual object, a minimal underlying temporal graph, or more formally the transitive reduction of the graph of relations between event boundaries [15] .