Combining Specialized Entailment Engines for RTE-4

The main goal of FBK-irst participation at RTE-4 was to experiment the use of combined specialized entailment engines, each addressing a specific phenomena relevant to entailment. The approach is motivated since textual entailment is due to the combination of several linguistic phenomena which interact among them in a quite complex way. We were driven by the following two considerations: (i) devise a general framework, based on distance between T and H, flexible enough to allow the combination of single entailment engines; (ii) provide a modular approach through which evaluate progresses on single aspects of entailment, using specialised traning and test dataset. For RTE4 we used two simple entailment engines, one addressing negation and the other lexical similarity, with a linear combination of their respective distances on T-H pairs.