Automatic Generation of Large-scale Paraphrases

Research on paraphrase has mostly fo-cussed on lexical or syntactic variation within individual sentences. Our con-cern is with larger-scale paraphrases, from multiple sentences or paragraphs to entire documents. In this paper we address the problem of generating paraphrases of large chunks of texts. We ground our discussion through a worked example of extending an exist-ing NLG system to accept as input a source text, and to generate a range of fluent semantically-equivalent alterna-tives, varying not only at the lexical and syntactic levels, but also in document structure and layout.