ABSTRACT
This chapter will describe specific ways in which technology is used in the work of the Patent Cooperation Treaty (PCT) Translation Division of the World Intellectual Property Organization (WIPO). After a brief introduction to the structure and work of the PCT Translation Division, the chapter will discuss the historical trajectory of the usage of technology in the Division, focusing on the areas of corpora, terminology, translation memories (TMs) and machine translation (MT). The workflows followed for translation and terminology work will be detailed, as well as the current applications of these tools. In particular, benefits and challenges of using TMs for Japanese into English translation of patent abstracts will be discussed. Subsequently, extraction of terms from monolingual and bilingual patent corpora will be described, as well as the use of term extraction and MT to optimize compilation of multilingual terminology records in the PCT terminology database. WIPO Pearl will be then introduced as a successful example of integration of corpora, MT, and innovative machine learning algorithms. To conclude, application of emerging technologies, i.e. neural machine translation (NMT), in the PCT Translation Division will be discussed with respect to the need of maintaining high quality levels of translation while fostering productivity.
