Complexity of Event Structure in IE Scenarios

This paper presents new Information Extraction scenarios which are linguistically and structurally more challenging than the traditional MUC scenarios. Traditional views on event structure and template design are not adequate for the more complex scenarios.The focus of this paper is to show the complexity of the scenarios, and propose a way to recover the structure of the event. First we identify two structural factors that contribute to the complexity of scenarios: the scattering of events in text, and inclusion relationships between events. These factors cause difficulty in representing the facts in an unambiguous way. Then we propose a modular, hierarchical representation where the information is split in atomic units represented by templates, and where the inclusion relationships between the units are indicated by links. Lastly, we discuss how we may recover this representation from text, with the help of linguistic cues linking the events.