关键词: 关键字提取, 小世界网络, 机会

Abstract: With the development of Internet，the event from large amount of Internet information，which is important and will become chance，can be extracted.The traditional algorithms，which are based on the model of high frequency，are hard to discover the important and unobserved event.In this paper，a new automatic keyword exaction method based on Small-World Network theory，is proposed.This method is based on the ideas of KeyGraph，constructs the graph of terms and relation，and is optimized by Small-World Network theory.The extracted words can represent more information of the document content，and the low frequent but important terms can also be extracted.Finally，through experiment validated，latency event，which is extracted by this method，is chance.

Key words: keyword extraction, small-world network, chance