Mutual implication as a measure of textual equivalence

Animesh Nighojkar, John Licato

Abstract: Semantic Textual Similarity (STS) and paraphrase detection are two NLP tasks that have a high focus on the meaning of sentences, and current research in both re-lies heavily on comparing fragments of text. Little to no work has been done in studying inference-centric approaches to solve these tasks. We study the relation be-tween existing work and what we call mutual implication (MI), a binary relationship between two sentences that holds when they textually entail each other. MI thus shifts the focus of STS and paraphrase detection to understanding the meaning of a sentence in terms of its inferential properties. We study the comparison between MI, paraphrasing, and STS work. We then argue that MI should be considered a complementary evaluation metric for advancing work in areas as diverse as machine translation, natural language inference, etc. Finally, we study the limitations of MI and discuss possibilities for overcoming them.