Sequence Labeling: A Practical Approach
This work provides a practical solution for sequence labeling in resource-scarce settings, though it is incremental as it builds on existing Bi-LSTM methods.
The authors tackled the sequence labeling problem under constraints of limited domain expertise and resources by developing a universal Bi-LSTM-based neural model that integrates morphological, semantic, and structural cues, achieving state-of-the-art results on four out of eight benchmark datasets across multiple NLP tasks and languages.
We take a practical approach to solving sequence labeling problem assuming unavailability of domain expertise and scarcity of informational and computational resources. To this end, we utilize a universal end-to-end Bi-LSTM-based neural sequence labeling model applicable to a wide range of NLP tasks and languages. The model combines morphological, semantic, and structural cues extracted from data to arrive at informed predictions. The model's performance is evaluated on eight benchmark datasets (covering three tasks: POS-tagging, NER, and Chunking, and four languages: English, German, Dutch, and Spanish). We observe state-of-the-art results on four of them: CoNLL-2012 (English NER), CoNLL-2002 (Dutch NER), GermEval 2014 (German NER), Tiger Corpus (German POS-tagging), and competitive performance on the rest.