This folder contains the output of our system GETARUNS for the Shared Task.
We created a folder for each text where we put one file for each sentence, and an additional file called DMWE, that is Discourse Model World of Entities, which is the input to the Question/Answering module of our system. This file contains all entities recorded with a semantic Identifier by the system during the analysis of the text. The file is produced by a procedure that recursively searches the dynamic database of FACTS or Infons in Situation Semantics terms, associated to each entity semantic identifier. These Infons may register properties, attributes or participation in events. Eventually, Infons may also be inherited in case one of the entity is semantically included in another entity - see the case of CANCER being included in the more general notion of CANCERS at the end of text 2.
The procedure produces a score that is derived from the relevance in terms of topichood - being Main, Secondary or Potential Topic - as asserted by the Centering algorithm. Entities and their associated infons are thus graded according to relevance.
They are listed on the basis of their ontological status: INDividuals, SETs, CLASSes.

Eventually, every Text folder may contain one Question file or a folder of Questions which have been formulated and answered by the system on the basis of the DMWE.