Aspect Detection using Word and Char Embeddings with (Bi)LSTM and CRF
This work improves aspect detection for sentiment analysis, but it is incremental as it builds on existing embedding and sequence modeling techniques.
The authors tackled aspect extraction in text by combining word and character embeddings with BiLSTM and CRF models, achieving state-of-the-art F-scores of 85% on SemEval Restaurants and 80% on Laptops datasets.
We proposed a~new accurate aspect extraction method that makes use of both word and character-based embeddings. We have conducted experiments of various models of aspect extraction using LSTM and BiLSTM including CRF enhancement on five different pre-trained word embeddings extended with character embeddings. The results revealed that BiLSTM outperforms regular LSTM, but also word embedding coverage in train and test sets profoundly impacted aspect detection performance. Moreover, the additional CRF layer consistently improves the results across different models and text embeddings. Summing up, we obtained state-of-the-art F-score results for SemEval Restaurants (85%) and Laptops (80%).