摘要
研讨系统会产生大量的专家发言文本信息,对发言文本进行聚类分析并可视化展现聚类结果,可以引导专家思维,提高研讨效率。提出一种基于聚类的研讨文本分析及可视化方法。首先对专家发言文本进行分词,提取专家发言文本向量;再采用聚类算法对专家发言文本进行聚类分析,得到不同发言簇;最后在基于Web的研讨系统中实现研讨文本聚类,并采用D3方法对聚类结果进行可视化展现。实验结果表明,该方法能有效展现专家发言之间的相似性和聚集度,引导专家思维收敛并达成共识。
The web-based argumentation support system（WASS） will produce a large number of experts＇ speech texts. Clustering analysis for speech texts and visualizing the clustering results can not only guide experts＇ thinking but also im prove the efficiency of argumentation. In this paper, a method of text clustering and visualization for WASS is proposed. First, the speeches text is segmented and the text vector is extracted. Then clustering algorithm is used to cluster the ex- perts＇ speeches and get different text dusters. Finally, the proposed method is realized in WASS, and the D3-based meth- od is used to visualize the clustering results. The experimental results show that the method can achieve clustering and vi- sualization of speech texts, thus improving the efficiency of the argumentation.
出处
《软件导刊》
2017年第4期14-17,共4页
Software Guide
基金
2016年度湖北省大学生创新创业训练计划项目(201610500030)
关键词
研讨支持系统
发言文本
文本聚类
可视化
Web-Based Argumentation Support System（WASS）
Speech Text
Text （
lustering
Visualization
