A Cross Language Text Categorization Algorithm from the Perspective of Information Retrieval

In this paper, we propose a novel method that performs Cross Language Text Categorization (CLTC) from the perspective of Information Retrieval. We present an input document in target language in the form of a query in source language. Then we retrieve the training documents in source language and find K most relevant results. At last, we use the class labels of the K results to predict the class of the input document. The only external resource required by our method is a bilingual dictionary. Experimental results show that our method gives promising performance, which is better than translation-based method.