Web page and image semi-supervised classification with heterogeneous information fusion

Web page and image semi-supervised classification with heterogeneous information fusion
复制标题

异构信息融合的网页和图像半监督分类

DOI:
10.1177/0165551513477818
复制
发表时间:
2013-06
影响因子:
2.4
通讯作者:
Xiaohong Guan
Xiaohong Guan
中科院分区:
计算机科学3区
文献类型:
--
作者:
Youtian Du;Chang Su;Zhongmin Cai;Xiaohong Guan

文献摘要

参考文献

相似文献

Web数据,例如网页和Web图像,可以自然地划分为多个异构属性集。具体来说,网页由超链接和内容组成,网页图像由文本和视觉信息组成。在本文中,我们提出了一种新的多视图半监督学习方法,称为局部协同训练,用于网页和图像分类。局部协同训练采用局部线性模型来表示每个视图(即一个属性集)上的数据点,并使用未标记的数据和协同训练策略迭代地细化它们。在每次迭代中,只有一部分局部模型(我们称之为主导局部模型)需要增量更新。因此,该方法非常高效,适合大规模网络数据的学习。此外,我们引入了一种基于置信度和分歧的新测量方法,以描述哪些未标记的示例对于丰富训练集“有好处”。局部协同训练在两种主要类型的半监督方法之间架起了一座桥梁:基于图的方法和协同训练。在网页和网络图像数据集上的实验表明,局部协同训练可以通过利用多个属性集和未标记数据来有效提高分类性能。
Web data, such as web pages and web images, can be naturally partitioned into multiple heterogeneous attribute sets. Concretely speaking, web pages consist of hyperlink and contents, and web images consist of the textual and visual information. In this paper, we propose a new multi-view semi-supervised learning method, named local co-training, for web page and image classification. Local co-training employs local linear models to represent data points on each view (i.e. one attribute set), and iteratively refines them using unlabelled data with co-training strategy. In each iteration, only a part of local models that we call dominant local models needs to be incrementally updated. The method is thus efficient and fit for the learning of large-scale web data. In addition, we introduce a new measurement based on both the confidence and the disagreement to describe which unlabelled examples are ‘good’ for the enrichment of training sets. Local co-training builds a bridge between two dominant types of semi-supervised methods: graph-based methods and co-training. Experiments on web page and web image datasets demonstrate that local co-training can effectively improve the classification performance by exploiting multiple attribute sets and unlabelled data.
DOI: 10.1109/tkde.2005.186
发表时间: 2005-11
影响因子: 8.9
作者:
Zhi-Hua Zhou;Ming Li
通讯作者: Zhi-Hua Zhou;Ming Li
DOI: 10.1109/tkde.2008.150
发表时间: 2009-02
影响因子: 8.9
作者:
Tao Jiang;A. Tan
通讯作者: Tao Jiang;A. Tan
DOI: --
发表时间: 2008-07
期刊: --
影响因子: --
作者:
Dan Zhang;Fei Wang;Changshui Zhang;Ta-Hsin Li
通讯作者: Dan Zhang;Fei Wang;Changshui Zhang;Ta-Hsin Li
DOI: --
发表时间: 2005
期刊: --
影响因子: --
作者:
Vikas Sindhwani;P. Niyogi;M. Belkin
通讯作者: Vikas Sindhwani;P. Niyogi;M. Belkin
DOI: 10.1109/tsmcb.2011.2157998
发表时间: 2011-12
期刊: IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics)
影响因子: --
作者:
Min-Ling Zhang;Zhi-Hua Zhou
通讯作者: Min-Ling Zhang;Zhi-Hua Zhou