Keyphrase Extraction for N-best Reranking in Multi-Sentence Compression

Florian Boudin and Emmanuel Morin

Multi-Sentence Compression (MSC) is the task of generating a short single sentence summary from a cluster of related sentences. This paper presents an N-best reranking method based on keyphrase extraction. Compression candidates generated by a word graph-based MSC approach are reranked according to the number and relevance of keyphrases they contain. Both manual and automatic evaluations were performed using a dataset made of clusters of newswire sentences. Results show that the proposed method significantly improves the informativity of the generated compressions.