同义词林.pptVIP

Corpus Annotation Shih-jui Lin 2003/07/22 Outline Corpus annotation (semi-automated approach) Word segmentation Part-of-speech tagging Bracketing (constructing parsing trees) Resources Corpus Tagged corpus Tree banks Bilingual parallel corpus Lexicon, thesaurus, ontology WordNet, 同義詞詞林, HowNet, InfoMap Goal For Linguistics To obtain syntactical and semantic knowledge about languages For CS/IT To construct training corpus for NLP research and automate the issues, such as QA, summarization, etc. Annotation Workflow Semi-automated Approach Word Segmentation Wordhood tests Frequency of co-occurre