ADMit: Improving NER in automotive domain with domain adversarial training and multi-task learning

Cheoneum Park, Seohyeong Jeong, Juae Kim

Abstract: Highlights•We construct the automotive NER dataset in both Korean and English.•We propose ADMit that utilizes domain adversarial and multi-task learning for NER.•We show that ADMit achieves state-of-the-art in the automotive NER datasets.•ADMit improves the performance of automotive FAQ system for real user’s queries.