首页>
外国专利>
Information processing apparatus, method, and program for obtaining weight for each feature amount in subjective hierarchical clustering
Information processing apparatus, method, and program for obtaining weight for each feature amount in subjective hierarchical clustering
展开▼
机译:用于在主观层次聚类中获得每个特征量的权重的信息处理设备,方法和程序
展开▼
页面导航
摘要
著录项
相似文献
摘要
The weight of each physical feature is learned so that hierarchical clustering reflecting the subjective similarity can be performed. The information processing apparatus of the present invention indicates a pair instructed by a user that a plurality of contents grouped by three (hereinafter referred to as “triplet contents”) have the highest degree of similarity between the three group contents. A plurality of pieces of learning data are acquired together with the label information. The information processing apparatus performs hierarchical clustering using the feature amount vector of each content of the learning data and the weight for each feature amount to obtain the hierarchical structure of the learning data. The information processing apparatus has a degree of matching between a pair that is first combined as being the same cluster among the triplet content in the obtained hierarchical structure and a pair indicated by the label information corresponding to the triplet content. The weight for each feature amount is updated so as to increase.
展开▼