Textual Paraphrase Dataset for Deep Language Modelling

Abstract The Turku Paraphrase Corpus is a dataset of over 100,000 Finnish paraphrase pairs. During the corpus creation, we strived to gather challenging pairs, more suitable test capabilities natural language understanding models. paraphrases are both selected and classified manually, so as minimise lexical overlap, provide examples that structurally lexically different maximum extent. An important distinguishing feature most pairs extracted distributed in their native document context, rather than isolation. primary application for development evaluation deep models, representation learning general.