Apple sponsored the Empirical Methods in Natural Language Processing (EMNLP) conference, which was held virtually from November 16 to 20. EMNLP is a leading conference focused on natural language processing.
Conference Accepted Paper
Jianpeng Cheng, Devang Agrawal, Hector Martinez Alonso, Shruti Bhargava, Joris Driesen Federico Flego, Dain Kaplan, Dimitri Kartsaklis, Lin Li, Dhivya Piraviperumal, Jason D Williams, Hong Yu, Diarmuid O Seaghdha, Anders Johannsen
We consider a new perspective on dialog state tracking (DST), the task of estimating a user's goal through the course of a dialog. By formulating DST as a semantic parsing task over hierarchical representations, we can incorporate semantic compositionality, cross-domain knowledge sharing and co-reference. We present TreeDST, a dataset of 27 thousand conversations annotated with tree-structured dialog states and system acts. We describe an encoder-decoder framework for DST with hierarchical representations, which leads to 20 percent improvement over state-of-the-art DST approaches that operate on a flat meaning space of slot-value pairs.
Conference Findings Accepted Paper
Shayne Longpre, Yu Wang, Christopher DuBois
Task-agnostic forms of data augmentation have proven widely effective in computer vision, even on pretrained models. In NLP similar results are reported most commonly for low data regimes, non-pretrained models, or situationally for pretrained models. In this paper we ask how effective these techniques really are when applied to pretrained transformers. Using two popular varieties of task-agnostic data augmentation (not tailored to any particular task), Easy Data Augmentation (Wei and Zou, 2019) and Back-Translation (Sennrichet al., 2015), we conduct a systematic examination of their effects across 5 classification tasks, 6 datasets, and 3 variants of modern pre-trained transformers, including BERT, XLNet, and RoBERTa. We observe a negative result, finding that techniques which previously reported strong improvements for not pre-trained models fail to consistently improve performance for pre-trained transformers, even when training data is limited. We hope this empirical analysis helps inform practitioners where data augmentation techniques may confer improvements.
Talks and Workshops
This workshop, was held on November 19, brought together Natural Language Processing (NLP) researchers, and Web Search and Information Retrieval specialists to collaborate on search-oriented conversational AI. We presented our accepted paper discussing the dependency between question formulation and correct answer selection in conversational question answering.
SCAI Workshop Accepted Paper:
A Wrong Answer or a Wrong Question? An Intricate Relationship between Question Reformulation and Answer Selection in Conversational Question Answering
Svitlana Vakulenko, Shayne Longpre, Zhucheng Tu, Raviteja Anantha
This workshop, held on November 20, was an opportunity to encourage development of more efficient NLP models and provide more simple architectures and empirical justification of model complexity. We presented our accepted paper which discussed the optimal combination of known techniques to optimize inference speed without sacrificing translation quality.
SustaiNLP Workshop Accepted Paper:
Efficient Inference for Neural Machine Translation
Yi-Te Hsu, Sarthak Garg, Yi-Hsiu Liao, Ilya Chatsviorkin