Sentence compression beyond word deletion
Trevor Cohn, Mirella Lapata · 2008
In this paper we generalise the sentence compression task.Rather than simply shorten a sentence by deleting words or constituents, as in previous work, we rewrite it using additional operations such as substitution, reordering, and insertion.We present a new corpus that is suited to our task and a discriminative tree-totree transduction model that can naturally account for structural and lexical mismatches.The model incorporates a novel grammar extraction method, uses a language model for coherent output, and can be easily tuned to a wide range of compression specific loss functions.O: Kurtz came from Missouri, and at the age of 14, hitchhiked to Los Angeles seeking top diving coaches.E: Kurtz came from Missouri, and at 14, hitch-hiked to Los Angeles seeking top diving coaches.A: Kurtz hitch-hiked to Los Angeles seeking top diving coaches.G: Kurtz came from Missouri, and at 14, hitch-hiked to Los Angeles seeking diving coaches.O: The scheme was intended for people of poor