摘要
针对生成文摘内容不完整的问题,利用相邻词的共现频率进行未登录词识别,提出一种通过词汇链的构建进行中文关键词抽取和文摘生成的算法,并给出一种采用《知网》为知识库构建词汇链的方法。通过计算词义相似度构建词汇链,结合词汇所在词汇链的强度、信息熵和出现位置等属性,进行关键词抽取和句子重要度计算。实验结果表明,与已有算法相比,该算法能够提高生成摘要的召回率和准确率。
In order to over the shorlcoming of the incomprehensive of summarization, a new lexical chain-based keywords extraction and automatic summarization algorithm from Chinese texts based on the unknown worst recognition using co-occurrence of neighbor words is proposed, and an algorithm for constructing lexical chain based on Hownet knowledge database is given in the method, lexical chain is constructed by calculating the semantic similarity between terms, keywords are extracted and the importance of each sentence is calculated according to the intensity of lexical chain, the entropy of terms and position. Experimental results show that the summarization generated by the improved algorithm gets better performance than other methods both in recall and precision.
出处
《计算机工程》
CAS
CSCD
2012年第3期183-186,共4页
Computer Engineering
基金
北京市优秀人才培养资助专项科研基金资助项目(2009D005001000005)
关键词
自动文摘
向量空间模型
关键词抽取
词汇链
未登录词识别
automatic summarization
vector space model
keyword extraction
lexical chain
unknown word recognition