摘要
提出一种基于相似融合的文本特征降维方法。首先求出不同聚类结果中每个簇的对应关系,然后计算这几个聚类结果对应簇的交集,对求交集后剩余的特征使用一种改进的互信息方法进行二次聚类,在最大限度减少信息损失的前提下实现了文本特征的有效降维。对文本的分类实验结果表明,该方法具有良好的降维效果,并且提高了聚类的效率。
A text feature reduction method based on similar combination is proposed in the paper. First the correlations among the different clusters are obtained and then the intersections of corresponding clusters are calculated. The remain features are clustered again with the improved mutual information method to reduce text features in the conditions of the least information loss. The experiment shows that this method has a good effect on dimensionality reduction and can improve the clustering ability.
出处
《长春工业大学学报》
CAS
2009年第6期651-656,共6页
Journal of Changchun University of Technology
基金
国家自然科学基金资助项目(60573182
69883004)
教育部博士点基金资助项目(20060183042)
吉林省科技发展计划项目(20060527
20040531)
关键词
文本分类
蚁群算法
互信息
相似融合
特征降维
text categorization
ant colony algorithm
mutual information
similar combination
feature reduction.