基于论文文本大数据挖掘的评价因子取值方法探索——以历史文化村镇保护评价为例

以历史文化村镇保护为案例,研究以论文文本大数据为数据源,采用文本挖掘技术和统计学原理进行评价因子取值的新方法。开放性的知网大数据可以为专家评价研究提供近似全数据样本和分析平台,可以提高主观评价的信度和效度。摸索出一套专家论文数据采集、清洗和去噪的方法,自主开发软件,以智能化的方式筛选出对评价因子研究有价值的专业词汇,并进行归类整理,得出历史文化村镇保护专家评价表及热点走势分析。

Exploration of the Evaluation Factor Selection Method based on Paper Text Big Data Mining: An Example of Historic Township Protection Evaluation

With historical and cultural village protection as the focus, this research takes big data from research texts as the data source and uses text mining technology and statistical principles as a new method for examining evaluation factors. Open big data from CNKI  can approximate a full data sample and is an analysis platform for expert evaluation research; this can improve the reliability and validity of subjective evaluation. A set of methods for data collection, cleaning, and de-noising of expert papers is explored in this paper. The independently developed software is able to intelligently screen out professional vocabularies with research reference to evaluation factors . Afterwards, the vocabularies are concluded and summarized, so as to obtain an evaluation table of historic and cultural protection experts and a trend analysis of hot spots.