Neural Entity Linking on Technical Service Tickets
This addresses entity linking for business applications with domain-specific terminology, but it is incremental as it applies an existing method to a new domain.
The paper tackled entity linking on technical service tickets, a practical business use case with scarce labels and low-quality text, and found that a neural approach based on BERT outperformed hand-coded heuristics by about 20% top-1 accuracy.
Entity linking, the task of mapping textual mentions to known entities, has recently been tackled using contextualized neural networks. We address the question whether these results -- reported for large, high-quality datasets such as Wikipedia -- transfer to practical business use cases, where labels are scarce, text is low-quality, and terminology is highly domain-specific. Using an entity linking model based on BERT, a popular transformer network in natural language processing, we show that a neural approach outperforms and complements hand-coded heuristics, with improvements of about 20% top-1 accuracy. Also, the benefits of transfer learning on a large corpus are demonstrated, while fine-tuning proves difficult. Finally, we compare different BERT-based architectures and show that a simple sentence-wise encoding (Bi-Encoder) offers a fast yet efficient search in practice.