Named entities identification

Apply

Named entities identification

Journal Title: Romanian Journal of Human - Computer Interaction - Year 2014, Vol 7, Issue 4

Abstract

An important topic in natural language processing is represented by named entities recognition inside texts. This article describes a novel approach used for detecting named entities that tries to improve the results obtained with the named entity recognition module from Stanford NLP library. In order to determine and classify named entities, this new model uses the Naive Bayes classifier. Our method is focused on named entities of type person and organization but it can be easily extended to other types of named entities. As training data we are using text that is manually annotated, text annotated with Stanford NLP toolkit and a set of XML files containing rules that describe different patterns. After the training, we are using the naive Bayes classifier in order to classify new entities. As test data we are using a Reuters collection of approximate 25000 articles among which 150 articles were manually annotated and used as training data. In order to evaluate the method we are computing the precision, the recall and the F1 factor.

Authors and Affiliations

Liviu Sebastian Matei ,Ştefan Trăuşan-Matu

Keywords

named entities Bayes machine learning natural language processing

EP ID EP28955
DOI -
Views 333
Downloads 8