作者: Dimitra Farmakiotou , John Koutsias , Panagiotis Stamatopoulos , Vangelis Karkaletsis , Constantine D. Spyropoulos
DOI:
关键词:
摘要: The identification and classification of proper names (named entity recognition) is considered an important task in the area Information Retrieval Extraction. A typical named recognition (NER) system mainly consists a lexicon grammar. When moving to new domain, these lexical resources should be customised, either manually or exploiting machine learning techniques. In this paper, we present NER based on hand crafted resources. part Greek information extraction was tested corpus financial news with satisfactory results.