Named Entity Recognition Architecture Combining Contextual and Global Features

Named entity recognition (NER) is an information extraction technique that aims to locate and classify named entities (e.g., organizations, locations, ...) within a document into predefined categories. Correctly identifying these phrases plays significant role in simplifying access. However, it remains difficult task because (NEs) have multiple forms they are context dependent. While the can be represented by contextual features, global relations often misrepresented those models. In this paper, we propose combination of features from XLNet Graph Convolution Network (GCN) enhance NER performance. Experiments over widely-used dataset, CoNLL 2003, show benefits our strategy, with results competitive state art (SOTA).