# An Effective Domain Adaptive Post-Training Method for BERT in Response Selection

## TL;DR

This paper introduces a domain-specific post-training method for BERT that significantly improves response selection in dialog systems by enhancing domain adaptation, achieving state-of-the-art results on benchmark datasets.

## Contribution

The paper proposes a highly effective post-training approach on domain-specific corpora to enhance BERT's performance in response selection tasks.

## Key findings

- Achieves 5.9% improvement on Ubuntu Corpus V1
- Achieves 6% improvement on Advising Corpus
- Sets new state-of-the-art performance on both benchmarks

## Abstract

We focus on multi-turn response selection in a retrieval-based dialog system. In this paper, we utilize the powerful pre-trained language model Bi-directional Encoder Representations from Transformer (BERT) for a multi-turn dialog system and propose a highly effective post-training method on domain-specific corpus. Although BERT is easily adopted to various NLP tasks and outperforms previous baselines of each task, it still has limitations if a task corpus is too focused on a certain domain. Post-training on domain-specific corpus (e.g., Ubuntu Corpus) helps the model to train contextualized representations and words that do not appear in general corpus (e.g., English Wikipedia). Experimental results show that our approach achieves new state-of-the-art on two response selection benchmarks (i.e., Ubuntu Corpus V1, Advising Corpus) performance improvement by 5.9% and 6% on R@1.