| 中文问答系统中答案抽取的研究 |
| |
| 引用本文： | 刘宁锋,史晓东. 中文问答系统中答案抽取的研究[J]. 数字社区&智能家居, 2011, 0(12): 2865-2868 |
| |
| 作者姓名： | 刘宁锋 史晓东 |
| |
| 作者单位： | 厦门大学; |
| |
| 摘 要： | 答案抽取是开放领域问答系统中的关健部分,主要由检索结果句子的重排和最终候选实体的抽取构成.通过把依存分析到引入到这两个步骤中,并结合改进的关键词密度算法,增强了对句法和语义知识的利用和系统的鲁棒性,在实验中答案抽取的效果明显提高.
|
| 关 键 词： | 问答系统 答案抽取 依存分析 语义角色标注 |
Answer Extraction Research for Chinese Question Answering System |
| |
| LIU Ning-feng,SHI Xiao-dong. Answer Extraction Research for Chinese Question Answering System[J]. Digital Community & Smart Home, 2011, 0(12): 2865-2868 |
| |
| Authors: | LIU Ning-feng SHI Xiao-dong |
| |
| Affiliation: | LIU Ning-feng,SHI Xiao-dong(Xiamen University,Xiamen 361005,China) |
| |
| Abstract: | Answering extraction is the key component of open domain question answering system,which contains sentences reranking and final candidate entity extraction.Through introducing dependency parsing into these two procedures incorporating Key words density algorithms,the system uses more syntactic and semantic knowledge as well as the enhancement of the system's robust and at last get a significant improvement in the experiment. |
| |
| Keywords: | question answering system answer extraction dependency parsing semantic role labeling |
| 本文献已被 CNKI 等数据库收录！ |
|
