# ThisIsCompetition at SemEval-2019 Task 9: BERT is unstable for out-of-domain samples

## TL;DR

This paper presents JESSI, a system combining multiple encoders including BERT for suggestion mining, highlighting BERT's instability on out-of-domain data and proposing methods to improve robustness.

## Contribution

The paper introduces a hybrid encoder system with domain adversarial training to enhance BERT's stability for out-of-domain samples in suggestion mining.

## Key findings

- BERT performs well in-domain but is unstable out-of-domain.
- Combining BERT with non-BERT encoders improves stability.
- The system achieved second place with high F-Score without external data.

## Abstract

This paper describes our system, Joint Encoders for Stable Suggestion Inference (JESSI), for the SemEval 2019 Task 9: Suggestion Mining from Online Reviews and Forums. JESSI is a combination of two sentence encoders: (a) one using multiple pre-trained word embeddings learned from log-bilinear regression (GloVe) and translation (CoVe) models, and (b) one on top of word encodings from a pre-trained deep bidirectional transformer (BERT). We include a domain adversarial training module when training for out-of-domain samples. Our experiments show that while BERT performs exceptionally well for in-domain samples, several runs of the model show that it is unstable for out-of-domain samples. The problem is mitigated tremendously by (1) combining BERT with a non-BERT encoder, and (2) using an RNN-based classifier on top of BERT. Our final models obtained second place with 77.78\% F-Score on Subtask A (i.e. in-domain) and achieved an F-Score of 79.59\% on Subtask B (i.e. out-of-domain), even without using any additional external data.