一种基于动态规划的缩写词定义识别方法

A dynamic programming algorithm for identifying definitions of abbreviations

生物文献挖掘是数据挖掘中的热点问题,论文针对文献挖掘中的缩写词定义识别问题提出了一种新的基于动态规划的比对算法,弥补了已有算法只能识别缩写词中的所有字符都来自于定义中字符这种形式的不足.实验结果表明,该算法相对于已有的缩写词定义识别算法取得了较好的回收率和准确率.

Biological text mining is a hot spot in data mining. In this paper, we proposed a new alignment algorithm for abbreviation-definition recognition in biological text mining based on dynamic programming. The other existed algorithms for abbreviation-definition recognition could only extract the abbreviation whose characters were all from the characters of its corresponding definition, and our algorithm could recognize more abbreviations and definitions. Experimental results illustrated that our algorithm achieved higher precision and recall than others.