Automatic construction of parallel English-Chinese corpus for cross-language information retrieval