项目后学习:环境影响后评价报告文本挖掘与分析

Learning from Post-Project Review:Text Mining and Analysis of Post Environmental Impact Review

项目后评价报告包含了大量的容易被忽视的隐性经验知识,是组织学习的重要知识来源。提出了一种基于LDA主题挖掘模型的经验学习方法,运用该方法,以环境影响后评价报告为数据来源,R语言作为分析工具,对隐含在项目环境影响后评价报告中的关键经验知识进行挖掘和分析;通过对重要主题进行筛选来明确样本所讨论的内容,对文档分类以明确样本每篇文档中所重点研究的内容,发现不同文档间的相似点;通过关键主题词分析,得出6条解决环境问题的重要经验知识,即预先制定全面的应急措施、从根源入手解决环境问题、实时关注作业区动态、从项目设备入手降低环境影响、重视沟通、考虑位置因素等;通过绘制主题关系网络图,不仅可以清楚主题间的关系,进一步可以得出隐含类环境经验知识主要来源于与"事故"相关的事件和项目本身。

The post-project evaluation report,which contains a lot of tacit experience knowledge easily neglected,is an important knowledge source of organizational learning.An empirical learning method based on LDA topic mining model was proposed.The key empirical knowledge implied in the post-project environmental impact review report was explored using R language.Analysis of the important topics was conducted to clarify the focus of the whole sample,and the research of document classification was carried out to clarify the key research contents in each document of sample and to find similarities between different documents.Through the analysis of key topic words,six important experience knowledge were obtained to solve environmental problems,including comprehensive emergency measures,search of root causes,focus on real-time operation dynamics,attention to project equipment,increased communication,and consideration of location factors.Analysis of the inter-topic network diagram show that tacit environmental knowledge originates from accident-related events and the project itself.