Timezone: »
Dialogue understanding tasks often necessitate abundant annotated data to achieve good performance and that presents challenges in low-resource settings. To alleviate this barrier, we explore few-shot data augmentation for dialogue understanding by prompting large pre-trained language models and present a novel approach that iterates on augmentation quality by applying weakly-supervised filters.We evaluate our methods on the emotion and act classification tasks in DailyDialog and the intent classification task in Facebook Multilingual Task-Oriented Dialogue. Models fine-tuned on our augmented data mixed with few-shot ground truth data are able to approach or surpass existing state-of-the-art performance on both datasets. For DailyDialog specifically, using 10% of the ground truth data we outperform the current state-of-the-art model which uses 100% of the data.
Author Information
Maximillian Chen (Columbia University)
Maximillian Chen is a second-year PhD student at Columbia University. His research has spanned persuasive dialogue systems, low resource techniques for dialogue tasks, and computational social science. Prior to Columbia, he received bachelors degrees in Computer Science and Statistics from Cornell University, where his research focused on computational social science and applied statistics.
Alexandros Papangelis (Amazon)
Chenyang Tao (Amazon)
Andy Rosenbaum (Amazon)
Seokhwan Kim (Amazon Alexa AI)
Yang Liu (Laix)
Zhou Yu (Columbia University)
Dilek Hakkani-Tur (Amazon Alexa AI)
More from the Same Authors
-
2021 Spotlight: Supercharging Imbalanced Data Learning With Energy-based Contrastive Representation Transfer »
Junya Chen · Zidi Xiu · Benjamin Goldstein · Ricardo Henao · Lawrence Carin · Chenyang Tao -
2021 : Towards Textual Out-of-Domain Detection without any In-Domain Labels »
Di Jin · Shuyang Gao · Seokhwan Kim · Yang Liu · Dilek Hakkani-Tur -
2021 : Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems »
Di Jin · Shuyang Gao · Seokhwan Kim · Yang Liu · Dilek Hakkani-Tur -
2023 Poster: Alexa Arena: A User-Centric Interactive Platform for Embodied AI »
Qiaozi Gao · Govindarajan Thattai · Suhaila Shakiah · Xiaofeng Gao · Shreyas Pansare · Vasu Sharma · Gaurav Sukhatme · Hangjie Shi · Bofei Yang · Desheng Zhang · Lucy Hu · Karthika Arumugam · Shui Hu · Matthew Wen · Dinakar Guthy · Shunan Chung · Rohan Khanna · Osman Ipek · Leslie Ball · Kate Bland · Heather Rocker · Michael Johnston · Reza Ghanadan · Dilek Hakkani-Tur · Prem Natarajan -
2022 : Towards Credible Human Evaluation of Open-Domain Dialog Systems Using Interactive Setup »
Sijia Liu · Patrick Lange · Behnam Hedayatnia · Alexandros Papangelis · Di Jin · Andrew Wirth · Yang Liu · Dilek Hakkani-Tur -
2022 Poster: Tight Mutual Information Estimation With Contrastive Fenchel-Legendre Optimization »
Qing Guo · Junya Chen · Dong Wang · Yuewei Yang · Xinwei Deng · Jing Huang · Larry Carin · Fan Li · Chenyang Tao -
2021 : Towards Zero and Few-shot Knowledge-seeking Turn Detection in Task-orientated Dialogue Systems »
Di Jin · Shuyang Gao · Seokhwan Kim · Yang Liu · Dilek Hakkani-Tur -
2021 Poster: Supercharging Imbalanced Data Learning With Energy-based Contrastive Representation Transfer »
Junya Chen · Zidi Xiu · Benjamin Goldstein · Ricardo Henao · Lawrence Carin · Chenyang Tao -
2020 Poster: Reconsidering Generative Objectives For Counterfactual Reasoning »
Danni Lu · Chenyang Tao · Junya Chen · Fan Li · Feng Guo · Lawrence Carin -
2019 Workshop: The third Conversational AI workshop – today's practice and tomorrow's potential »
Alborz Geramifard · Jason Williams · Bill Byrne · Asli Celikyilmaz · Milica Gasic · Dilek Hakkani-Tur · Matt Henderson · Luis Lastras · Mari Ostendorf -
2019 Poster: Improving Textual Network Learning with Variational Homophilic Embeddings »
Wenlin Wang · Chenyang Tao · Zhe Gan · Guoyin Wang · Liqun Chen · Xinyuan Zhang · Ruiyi Zhang · Qian Yang · Ricardo Henao · Lawrence Carin -
2019 Poster: On Fenchel Mini-Max Learning »
Chenyang Tao · Liqun Chen · Shuyang Dai · Junya Chen · Ke Bai · Dong Wang · Jianfeng Feng · Wenlian Lu · Georgiy Bobashev · Lawrence Carin -
2018 Poster: Adversarial Text Generation via Feature-Mover's Distance »
Liqun Chen · Shuyang Dai · Chenyang Tao · Haichao Zhang · Zhe Gan · Dinghan Shen · Yizhe Zhang · Guoyin Wang · Dinghan Shen · Lawrence Carin -
2015 Workshop: Machine Learning for Spoken Language Understanding and Interactions »
Asli Celikyilmaz · Milica Gasic · Dilek Hakkani-Tur