基于碎片化UGC的知识元抽取研究 |
| |
引用本文: | 王忠义,郑鑫.基于碎片化UGC的知识元抽取研究[J].情报理论与实践,2021(1):188-194. |
| |
作者姓名: | 王忠义 郑鑫 |
| |
作者单位: | 华中师范大学信息管理学院 |
| |
基金项目: | 教育部人文社会科学研究青年基金项目“大数据环境下碎片化用户生成内容的多粒度知识组织研究”的成果,项目编号:19YJC870025。 |
| |
摘 要: | 目的/意义]在大数据环境下,从海量的碎片化用户生成内容中抽取具有完整语义的知识单元。方法/过程]文章提出一种基于碎片化UGC的知识元抽取方法,该方法首先借助BTM主题分割方法从UGC中抽取知识要素,而后基于融合TextRank和Glove词向量的K-means方法实现知识要素聚类,最后根据知识要素相关属性和知识要素聚类结果生成对应UGC知识元。结果/结论]实验结果显示基于碎片化UGC的知识元抽取方法具有一定科学性和有效性。
|
关 键 词: | 用户生成内容 知识元 知识元抽取 主题分割 碎片化 |
Research on Knowledge Element Extraction Based on Fragmented UGC |
| |
Abstract: | Purpose/significance] In big data environment,to extract a complete semantic knowledge unit,from the vast amounts of fragmented user generated content(UGC).Method/process] this paper proposes a method of knowledge element extraction based on fragmented UGC,the method first using the BTM topic-based segmentation method to extract the knowledge fragments from the UGC,and then based on the K-means clustering method that incorporates TextRank and Glove to realize knowledge fragments clustering.At last,according to the knowledge elements attributes and knowledge fragments clustering results generated UGC knowledge elements.Result/conclusion] Experimental results show that the knowledge element extraction method based on fragmented UGC is scientific and effective. |
| |
Keywords: | UGC knowledge element knowledge element extraction topic segmentation fragment |
本文献已被 维普 等数据库收录! |
|