Cross Language Text Classification by Model Translation and Semi-Supervised Learning

In this paper, we introduce a method that automatically builds text classifiers in a new language by training on already labeled data in another language. Our method transfers the classification knowledge across languages by translating the model features and by using an Expectation Maximization (EM) algorithm that naturally takes into account the ambiguity associated with the translation of a word. We further exploit the readily available unlabeled data in the target language via semisupervised learning, and adapt the translated model to better fit the data distribution of the target language.

• Type: Journal
• Year: 2010
• Where: EMNLP
• Authors: Lei Shi, Rada Mihalcea, Mingjun Tian
• Keywords: Available Unlabeled Data, EMNLP 2010, Expectation Maximization, Natural Language Processing, Target Language