An Analysis of Deep Contextual Word Embeddings and Neural Architectures for Toponym Mention Detection in Scientific Publications

摘要

Toponym detection in scientific papers is an open task and a key first step in place entity enrichment of documents. We examine three common neural architectures in NLP: 1) convolutional neural network, 2) multi-layer perceptron (both applied in a sliding window context) and 3) bi-directional LSTM and apply contextual and non-contextual word embedding layers to these models. We find that deep contextual word embeddings improve the performance of the bi-LSTM with CRF neural architecture achieving the best performance when multiple layers of deep contextual embeddings are concatenated. Our best performing model achieves an average Fl of 0.910 when evaluated on overlap macro exceeding previous state-of-the-art models in the toponym detection task.