Adobe at ACL 2022

May 24, 2022

In a Findings of ACL 2022 paper, Adobe Research and university collaborators present a comprehensive analysis of the global and local attention patterns in Transformer models and propose a fine-tuning method that goes beyond these patterns by learning an adaptive attention pattern tuned to a specific downstream task.

The Annual Meeting of the Association for Computational Linguistics (ACL 2022) is being held from May 22 to 27, 2022. ACL is one of the top research conferences on natural language processing. In recent years, deep learning approaches have been prominently featured in the papers presented at this conference. 

Adobe Research has co-authored a total of six papers at the conference, one paper at the main conference, four papers in the Findings category, and one workshop paper. The research topics range from named entity recognition, text generation, offensive text detection, event extraction, and many more.  

Findings papers, while not accepted for publication in the main conference, were assessed by the ACL 2022 Program Committee to contain solid work with sufficient substance, quality, and novelty to be included in the conference proceedings. 

Many of the accepted papers are the result of student internships or other collaborations with university students and faculty. Check out the Adobe Research Careers page to learn more about internships and full-time career opportunities

ACL 2022 conference – Adobe co-authored papers 

Main conference paper 

Few-Shot Class-Incremental Learning for Named Entity Recognition 
Rui Wang, Tong Yu, Handong Zhao, Sungchul Kim, Subrata Mitra, Ruiyi Zhang, Ricardo Henao 

Findings papers 

CaM-Gen: Causally-aware Guided Text Generation 
Navita Goyal, Roodram Paneri, Ayush Agarwal, Udit Kalani, Abhilasha Sancheti, Niyati Chhaya 

Document-Level Event Argument Extraction via Optimal Transport 
Amir Pouran Ben Veyseh, Minh Van Nguyen, Franck Dernoncourt, Bonan Min, Thien Huu Nguyen 

Learning Adaptive Axis Attentions in Fine-tuning:  Beyond Fixed Sparse Attention Patterns 
Zihan Wang, Jiuxiang Gu, Jason Kuen, Handong Zhao, Vlad I Morariu, Ruiyi Zhang, Ani Nenkova, Tong Sun, Jingbo Shang 

Transfer Learning and Prediction Consistency for Detecting Offensive Spans of Text 
Amir Oururan Ben Veyseh, Ning Xu, Quan Tran, Varun Manjunatha, Franck Dernoncourt, Thien Nguyen 

Workshop paper 

How does fake news use a thumbnail? CLIP-based Multimodal Detection on the Unrepresentative News Image 
Hyewon Choi, Yejun Yoon, Seunghyun Yoon, Kunwoo Park  
Presented at the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations (CONSTRAINT) 

Related Posts