一种利用近邻和信息熵的主动文本标注方法  

An Active Labeling Method for Text Data Based on Nearest Neighbor and Information Entropy

作　　者：朱岩[1] 景丽萍[1] 于剑[1] 

机构地区：[1]北京交通大学计算机科学系,北京100044

出　　处：《计算机研究与发展》2012年第6期1306-1312,共7页Journal of Computer Research and Development

基　　金：中央高校基金科研业务费专项资金项目(2009YJS026);北京交通大学优秀博士生科技创新基金项目(141097522);国家自然科学基金项目(0905028;90820013;60875031)

摘　　要：由于大规模标注文本数据费时费力,利用少量标注样本和大量未标注样本的半监督文本分类发展迅速.在半监督文本分类中,少量标注样本主要用来初始化分类模型,其合理性将影响最终分类模型的性能.为了使标注样本尽可能吻合原始数据的分布,提出一种避开选择已标注样本的K近邻来抽取下一组候选标注样本的方法,使得分布在不同区域的样本有更多的标注机会.在此基础上,为了获得更多的类别信息,在候选标注样本中选择信息熵最大的样本作为最终的标注样本.真实文本数据上的实验表明了提出方法的有效性.As it is quite time-consuming to label text documents on a large scale, a kind of text classification with a few labeled data is needed. Thus, semi-supervised text classification emerges and develops rapidly. Different from traditional classification, semi-supervised text classification only requires a small set of labeled data is us of labeled data and a large set of unlabeled data to ed to initialize the classification model in most case performance of the final classifier. In order to consistent with the distribution of the original dat the K nearest neighbors of the labeled data to be method, the data located in various regions will h order to obtain more eategory information from th information entropy of the candidate labeled data is chosen as the next datum to be labeled manual this approach is very effective. train a classifier. s. Its rationality The small set will affect the

关 键 词：半监督文本分类 主动学习 近邻 信息熵 标注方法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]