Towards Component-Based Textual Entailment

In the Textual Entailment community, a shared effort towards a deeper understanding of the core phenomena involved in textual inference is recently arose. To analyse how the common intuition that decomposing TE would allow a better comprehension of the problem from both a linguistic and a computational viewpoint, we propose a definition for strong component-based TE, where each component is in itself a complete TE system, able to address a TE task on a specific phenomenon in isolation. We review the literature according to our definition, trying to position relevant work as more or less close to our idea of strong component-based TE. Several dimensions of the problem are discussed: i) the implementation of system components to address specific inference types, ii) the analysis of the phenomena relevant to component-based TE, and iii) the development of evaluation methodologies to assess TE systems capabilities to address single phenomena in a pair.