Predicting Relevance and Quality of Machine Translation for Product Reviews

Funded by the Amazon Academic Research Awards (AARA) program, this project was to devise a Quality Estimation (QE) approach for the machine translation (MT) of product reviews. On online market platforms such as Amazon, product reviews are abundant but written in a single language (often English). Automatically translating such reviews could better enable products to reach foreign markets. However, this type of content introduces important challenges to state of the art machine translation, which often results in far from perfect quality translations, and thus automatic quality estimation becomes paramount.