Mining Historical Social Issues
Mining Historical Social Issues
复制标题
挖掘历史社会问题
DOI：
10.1007/978-3-319-19857-6_50
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Ryohei IKEJIRI
中科院分区：
文献类型：
--
作者：
Yasunobu SUMIKAWA;Ryohei IKEJIRI
本文提出了一个框架，用于识别类似于学习者指定的现代社会问题的人类历史。从历史数据中，学习者可以研究历史上的人们是如何试图解决社会问题的，他们取得了什么结果。这可以帮助学习者思考如何解决现代社会问题。为了识别与给定现代问题相似的历史问题，我们的框架在两种技术中使用了特定现代问题的特征和解释：聚类和分类。这些技术通过使用矩阵运算和文本分类来识别历史问题和现代问题之间的相似性。我们实现了我们提出的框架，并从分析时间的角度对其进行了评估。实验结果表明，该框架具有实际应用价值，分析时间仅为S的0.7%左右。
This paper presents a framework for identifying human histories that are similar to a modern social issue specified by a learner. From the historical data, the learner can study how people in history tried to resolve social issues and what results they achieved. This can help the learner consider how to resolve the modern social issue. To identify issues in history similar to a given modern issue, our framework uses the characteristics and explanation of the specified modern issue in two techniques: clustering and classification. These techniques identify the similarity between historical and the modern issues by using matrix operation and text classification. We implemented our proposed framework and evaluated it in terms of analysis time. Experimental results proved that our framework has practical usage with an analysis time of only about 0.7 s.