语义增强的科技创新内容表征研究

Research on the Representation of Technical Innovation Content with Enhanced Semantics

[目的/意义]随着计算机信息处理技术以及文本数据挖掘技术的发展,研究人员开始利用语义分析技术深入分析科技文献文本数据,识别出科技文献中的科技创新内容,以期为科技创新和科技决策提供支持和帮助。[方法/过程]文章通过分析科技创新内容结构分布特征,以句子为最小标引粒度,利用Keygraph算法抽取出科技文献摘要中的关键词进行科技创新内容特征选择,基于SVM的语义角色标注技术完成科技创新内容的语义表征。[结果/结论]实验结果表明,语义增强的科技创新表征方法可以基本实现科技创新内容的语义标引。

[ Purpose/significance ] With the development of computer information processing technology and text mining tech- nology, researchers use semantic analysis method to analyze scientific literature data and identify technical innovation content in the scientific literature in order to provide support and assistance for technical innovation and technical decision-making. [ Method/ process ] By analyzing the distribution structure of technical innovation content, the paper takes sentence as the minimum indexing granularity, and uses Keygraph algorithm to extract the keywords in the abstract of scientific literature for the selection of character- istics of technical innovation content. Based on the semantic role labeling technique of SVM, the paper accomplishes the semantic representation of technical innovation content. [ Result/conclusion] The experimental results show that technical innovation repre- sentation method with enhanced semantic can basically achieve semantic indexing of technical innovation content.