基于关联规则的调查问卷多项选择题分析

Analysis of the Multiple-choice Questions in Questionnaires Based on the Relation Rule

为在调查问卷的数据分析中揭示隐藏在数据背后更本质的特点和深层规律,引入数据挖掘中关联规则分析的思想和计算方法,针对问卷调查的多项选择题,旨在解决同时频繁出现的若干选项,即频繁集的分析问题。由于问卷调查是基于抽样调查,所以需要解决以样本频繁集估计总体频繁集的问题。提出总体参数假设值的选取方法。以本研究的方法在某证券市场监管问题的研究中进行实证,取得良好效果。

In this paper, the authors address the issue of frequent itemsets in questionnaire analysis by introducing the idea of the relation rule analysis and methods of measuring, both of which are employed in data mining. The authors also present an approach to estimating population parameters in the presence of frequent itemsets. The methods introduced in this paper have been tested and proved effective in a capital market related research.