Abstract: Data mining technology has provided powerful support in realizing the value of big data, and text mining for data is one of its important aspects. The study of users' assessments is an important part of post-occupancy evaluation (POE) of built environments. As a result of abundant network information, many platforms have provided user evaluations including opinions relating to the built environment. This assessment information is extensive and can, following collection and analysis by text mining technology, be used to supplement traditional POE methods. This paper applies text mining technology to collect and analyze user opinions relating to the National Centre for the Performing Arts, the Shanghai Grand Theatre and the Guangzhou Opera House from network platforms, and verifies the feasibility of using this technology to supplement traditional PEO methods.

Key words: post-occupancy evaluation(POE), text mining, theatre, word frequency analysis

摘要： 数据挖掘技术为实现大数据理念的价值提供了有力的支持，针对于文字数据的文本挖掘技术是其重要的方向之一。在建成环境使用后评价中很重要的一部分是研究使用者对所处环境的评价意见，如今网络信息丰富，有众多平台提供用户评价服务，其中不乏使用者对建成环境的评价意见，这部分评价信息数量多、内容丰富，以文本挖掘技术对其进行采集和分析，可以作为传统使用后评价的补充。因此以国家大剧院、上海大剧院和广州大剧院作为实例，应用文本挖掘技术采集和分析使用者在网络平台上发表的对三座剧院的评价意见，验证该方法是否能对使用后评价传统方法做出补充。

关键词: 使用后评价, 文本挖掘, 剧院, 词频分析