Mining textual-feature based research on translator's styles

The utilization of Artificial Intelligence techniques in mining translation studies brings new discoveries on the research of translator's styles. Aided by the software WordSmith Tools and the Antconcordance, this paper conducts an empirical study of translator's styles by analyzing three English versions of the most famous Chinese love poem-Chang hen ge written by Bai Juyi in Tang Dynasty from three textual levels. The major experiment findings exhibit that the translator from the source language culture is prone to be paraphrase and translators from the target language culture tend to be metaphrase.