Computer-aided annotation of lexical cohesive devices in parallel texts for translator and interpreter training

Lexical cohesion, which concerns the relatedness of semantic meaning in a text, has been regarded as an indicator of one's linguistic competence. Our previous work has shown the use of lexical cohesion devices as an effective parameter for evaluation of machine translation. For translator and interpreter training, the use of lexical cohesive devices has also been shown to reveal a student translator's or interpreter's competence level. 

This chapter reports the ongoing work on the construction of the Chinese/English Translation and Interpreting Learner Corpus (CETILC) which features the annotation of lexical cohesive devices in the translation and interpretation outputs by students at various competence levels. We will illustrate a computer-aided annotation approach which combines automatic annotation and human verification with the aid of lexical resources, WordNet and Chinese Open WordNet, for supporting the annotation of various types of lexical cohesion devices in both English and Chinese texts. We will also present initial findings on the CETILC data for identifying the professional development path of student translators and interpreters through their use of lexical cohesion devices, which suggests the value of the corpus as a useful resource for translator and interpreter training.