Duration Modeling with Semi-Markov Conditional Random Fields for Keyphrase Extraction

Existing methods for keyphrase extraction need preprocessing to generate candidate phrase or post-processing transform keyword into keyphrase. In this paper, we propose a novel approach called duration modeling with semi-Markov Conditional Random Fields (DM-SMCRFs) extraction. First of all, based on the property chain, DM-SMCRFs can encode segment-level features and sequentially classify in sentence as non-keyphrase. Second, by assuming independence between state transition duration, model distribution (length) keyphrases further explore information, which help identify size Based convexity parametric feature derived from distribution, constrained Viterbi algorithm is improve performance decoding DM-SMCRFs. We thoroughly evaluate datasets various domains. The experimental results demonstrate effectiveness proposed model.