TL;DRAbstract
Named Entity Recognition (NEE) and baseNP chunking are two examples of NLP tasks that annotate chunks, i.e. non-embedding, non-overlapping token sequences. Many approaches to these tasks can be greatly simplified by viewing chunking as a tagging problem (Ramshaw and Marcus 1995). This perspective, however, raises the question of data representation, i. e. how chunk structures can be mapped onto tags, and whether the choice of chunk-tag mapping affects the system performance (cf. Tjong Kim Sang and Veenstra 1999).
Chat with Paper
AI Agents for this Paper
Named Entity Recognition (NEE) and baseNP chunking are two examples of NLP tasks that annotate chunks, i.e. non-embedding, non-overlapping token sequences. Many approaches to these tasks can be greatly simplified by viewing chunking as a tagging problem (Ramshaw and Marcus 1995). This perspective, however, raises the question of data representation, i. e. how chunk structures can be mapped onto tags, and whether the choice of chunk-tag mapping affects the system performance (cf. Tjong Kim Sang and Veenstra 1999).
Keywords
Chat
Click to start Chat