王伟钧, 敬灏, 苟娜, 杨晋浩. 基于多选问题数据规约的特征集分析方法研究[J]. 电子科技大学学报社科版, 2018, 20(1): 66-70. DOI: 10.14071/j.1008-8105(2017)-1003
引用本文: 王伟钧, 敬灏, 苟娜, 杨晋浩. 基于多选问题数据规约的特征集分析方法研究[J]. 电子科技大学学报社科版, 2018, 20(1): 66-70. DOI: 10.14071/j.1008-8105(2017)-1003
WANG Wei-jun, JING Hao, GOU Na, YANG Jin-hao. Research of Analysis Methods of Characteristics Set for Data Reduction on Multiple-choice Questions[J]. Journal of University of Electronic Science and Technology of China(SOCIAL SCIENCES EDITION), 2018, 20(1): 66-70. DOI: 10.14071/j.1008-8105(2017)-1003
Citation: WANG Wei-jun, JING Hao, GOU Na, YANG Jin-hao. Research of Analysis Methods of Characteristics Set for Data Reduction on Multiple-choice Questions[J]. Journal of University of Electronic Science and Technology of China(SOCIAL SCIENCES EDITION), 2018, 20(1): 66-70. DOI: 10.14071/j.1008-8105(2017)-1003

基于多选问题数据规约的特征集分析方法研究

Research of Analysis Methods of Characteristics Set for Data Reduction on Multiple-choice Questions

  • 摘要: 目前在调查分析领域,多选问题的分析主要针对各选项的分析和推断,而忽略了它们之间的关联信息。基于此提出了基于多选问题数据规约的特征集分析方法。首先,将多选问题的各选项组合分别看作一个特征组合,而所有的特征组合(即特征集)代表其对应的多选问题的全部信息,并对特征集进行数值编码;其次,为了解决特征集组合的“维度爆炸”问题,提出了累计比和聚类两种数据规约方法,不但保留了数据的主要信息,而且对规约后的维度给出很好的解释;最后,实际调查分析案例分析表明了该方法的合理性和有效性。该研究对于各领域关于多选问题的深度分析提供了技术支持。

     

    Abstract: At present, the analysis of multiple choice questions mainly focuses on the analysis and deduction of independent options in the field of survey and analysis without considering their relationships more or less. This paper presents two feature set analysis methods based on data reduction of multi-selection questions. Firstly, one or more options of the multiple-choice questions is considered as a feature combination, on behalf of its corresponding multiple choice of all the information, and then feature combination set (i.e. feature set) is digitally encoded. Secondly, in order to solve the problem of " dimensional explosion” of feature set, two kinds of data statistic methods, cumulative ratio and clustering, are proposed. Finally, the investigation of a case study shows the method’s rationality and effectiveness. The study provides a technical support for in-depth analysis of multiple selection issues in various fields.

     

/

返回文章
返回