基于SMA信息抽取的主题诊断研究

Study of Theme's Diagnosis Based on Information Extraction of SMA

从和谐主题的提出与研究现状入手,在建立策略性管理活动描述文本集合并从中抽取管理任务信息和管理问题信息的基础上,为了实现对管理领域中事实主题的识别与评价,本研究综合分析了从文本表示、维数约减到聚类分析以及数据概化的思想及其方法技术,并通过引入W正态分布检验,完成了对所研究问题的解决方案设计。实证分析结果验证了本方法对于解决组织管理中事实主题诊断问题的有效性和可行性。

From the viewpoint of putting forward and its investigative actuality of He-Xie theme, on the basis of foundation of descriptive text of SMA and information extraction of managerial tasks and managerial questions, for the sake of identifying and appraising the factual theme correctly in the domain of management, this research analyze in detail the thought and technical traits about text＇s expression, dimension＇s deduction, clustering analysis and data generalizability, and through the introduction of Shapiro-Wilk test of normal distribution, then the design of solution is realized. The empirical results indicate that this solution is both effectual and feasible to diagnosing of factual theme.