EVITA - Events In Text Analyzer

Evita is an event recognition system. It locates and tags all event-referring expressions in the input text that can be temporally ordered. This excludes generic events and states not temporally anchored. In addition, Evita identifies those grammatical features associated with the event-denoting expressions which inform about temporal and modal information of events; namely, tense, aspect, polarity, modality, as well as the event class. More details on Evita are available in the "Event and Time Tagging" section of the TimeML publications page.

Event annotation follows version 1.2.1 of the TimeML specifications. An example of event annotation is shown below, where the bold type indicates tags added by Evita.

In Washington today, the Federal Aviation Administration <EVENT eid="e1" class="OCCURRENCE">released</EVENT> air traffic control tapes from the night the TWA Flight eight hundred <EVENT eid="e2" class="OCCURRENCE">went</EVENT> down.

<MAKEINSTANCE eventID="e1" eiid="ei1" pos="VERB" tense="PAST" aspect="NONE"/>
<MAKEINSTANCE eventID="e2" eiid="ei2" pos="VERB" tense="PAST" aspect="NONE"/>

The input to Evita is chunked and part-of-speech tagged text (chunks and tags are not displayed in the example above).