Construction and Application of English Language Context-Driven Multimodal Corpus

This paper investigates the approach and process of constructing context-driven multimodal corpus with Elan software in terms of its diversity of annotation tiers, controlled vocabulary and speed control. By creating the rich and real language environment in diverse context and providing potentially user-customizable strategies, the construction of multimodal corpus for English learning represents an ideal paradigm of data-driven learning model. It is conducive to inspire the learner with great interest and enthusiasm in language learning, and thus improve their language proficiency and cognitive capability by extracting knowledge from corpus compelled by data and inducting abstract linguistic rules by locating authentic examples of language in real-world language use. Based on the multimedia data-driven concordancing, this paper further elaborates the feasible pedagogic application of multimodal corpus in such aspects as vocabulary differentiation, communication contextualization, listening perception, cross-culture awareness cultivation and multiliteracies teaching. This paper concludes with suggestions for surmounting the challengeable disadvantages that multimodal corpus entailed.