Towards a Visual Lexicon : The Creation of A Corpus Linguistic Database Using Digital Movie Data