Cross Language Text Classification via Subspace Co-regularized Multi-view Learning

By Yuhong Guo, Min Xiao. 2012

63 citations recorded by the source corpus.

In many multilingual text classification problems, the documents in different languages often share the same set of categories. To reduce the labeling cost of training a classification model for each individual language, it is important to transfer the label knowledge gained from one language to another language by conducting cross language classification. In this paper we develop a novel subspace co-regularized multi-view learning method for cross language text classification. This method is built on parallel corpora produced by machine translation. It jointly minimizes the training error of each classifier in each language while penalizing the distance between the subspace representations of parallel documents. Our empirical study on a large set of cross language text classification tasks shows the proposed method consistently outperforms a number of inductive methods, domain adaptation methods, and multi-view learning methods.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Saeid Motiian
2. Gianfranco Doretto
3. Tao Mei
4. Ting Yao
5. Yingwei Pan
6. Q. Jones
7. Seyed Mehdi Iranmanesh
8. D. Adjeroh

In Computer Vision and Pattern Recognition

1. Tao Mei
2. Ting Yao
3. Yingwei Pan
4. Luc Van Gool
5. Dacheng Tao
6. Gianfranco Doretto
7. Saeid Motiian
8. Yi Yang

Citations in the corpus

• Few-Shot Adversarial Domain Adaptation (2017, background): “Domain Adaptation can be supervised [58, 27], unsupervised [15, 34], or semi-supervised [16, 21, 63], depending on what data is available in a labeled format and how much can be collected.”
• Unified Deep Supervised Domain Adaptation and Generalization (2017, background): “Domain adaptation can be either supervised [60, 33], unsupervised [38, 38, 61], or semi-supervised [26, 29, 67].”
• Semi-supervised Domain Adaptation with Subspace Learning for visual recognition (2015, methodology): “The research on domain adaptation has proceeded along three different dimensions: unsupervised domain adaptation [1, 16, 25, 26], supervised domain adaptation [2, 8, 21, 24, 28], and semi-supervised domain adaptation [9, 13, 14].”