Named Entity Recognition for Norwegian
Named Entity Recognition (NER) is the task of identifying and classifying proper nouns in running text. The identified names are typically classified into categories like person, organizations, locations, events, etc. NER models typically combine several types of approaches, e.g., combining statistical machine learning methods with efficient look-up in large gazetteers (name lists). Although there currently is no off-the-shelf NER model available for Norwegian, we do have training data labeled with name tags. The aim of this project is to implement efficient and large-scale Named Entity Recognition for Norwegian text, wrapped in a portable tool that can be distributed for public use. As an example of an existing system that can re-trained on Norwegian data, see NeuroNER (based on using so-called word embedding representations as input to a neural network architecture for classification).