Debjit Paul

About Me

I'm a Postdoctoral Researcher working with Boi Faltings, Robert West, and Antoine Bosselut at EPFL. I did my Ph.D. under the supervision of Anette Frank in the Department of Computational Linguistics at Heidelberg University and also a part of Research Training Group AIPHES.

My primary research interest areas are : Natural Language Understanding, Commonsense Reasoning, Knowledge-Based Reasoning, Neuro-Symbolic NLP, Explainable Machine Learning, Deep Learning for NLP.

Email: debjit.paul (at) epfl.ch

News

Talk at KU Leuven: Neuro-Symbolic Commonsense Reasoning in NLP.

Successfully defended my PhD. Thesis (April 2022).

I will be joining School of Computer and Communication Sciences at EPFL as a PostDoc in April 2022.

I will be joining Amazon Alexa AI team, Berlin as Research Intern in October 2021.

Our paper COINS: Dynamically Generating COntextualized Inference Rules for Narrative Story Completion is accepted in ACL 2021.

Our paper Generating Hypothetical Events for Abductive Inference is accepted in StarSem 2021.

Our paper CO-NNECT: A Framework for Revealing Commonsense Knowledge Paths as Explicitations of Implicit Knowledge in Texts is accepted in IWCS 2021.

Giving an invited talk at Edinburgh Napier University, November 2020.

Our paper Social Commonsense Reasoning with Multi-Head Knowledge Attention was accepted in Findings of the Association for Computational Linguistics: EMNLP 2020.

Attending Computational Models of Argument (COMMA). September 2020.

Our paper Argumentative Relation Classification with Background Knowledge was accepted in COMMA 2020.

Attending Association for Computational Linguistics (ACL). July 2020.

Attending EurNLP 2019 in London, UK. October 2019.

Attending NAACL 2019 in Minneapolis, US. June 2019.

Our paper Handling Noisy Labels for Robustly Learning from Self-Training Data for Low-Resource Sequence Labeling was accepted in NAACL SRW 2019.

Our paper Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs was accepted in NAACL 2019.

Selected Projects

Handling Noisy Labels for Robustly Learning from Self-Training Data for Low-Resource Sequence Labeling

(NAACL-SRW) 2019 Read Paper Code

Argumentative Relation Classification with Background Knowledge

COMMA 2020 Read Paper Code Video

Selected Publications

Martin Josifoski, Maxime Peyrard, Frano Rajic, Jiheng Wei, Debjit Paul, Valentin Hartmann, Barun Patra, Vishrav Chaudhary, Emre Kıcıman, Boi Faltings, Robert West "Language Model Decoding as Likelihood-Utility Alignment" Findings @ EACL 2023

Debjit Paul, and Anette Frank "COINS: Dynamically Generating COntextualized Inference Rules for Narrative Story Completion" Proceedings of the Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (ACL-IJCNLP 2021), Online, Association for Computational Linguistics, long Paper, to appear. , 2021

Debjit Paul, and Anette Frank "Generating Hypothetical Events for Abductive Inference" Proceedings of The Tenth Joint Conference on Lexical and Computational Semantics (*SEM 2021), vol. Long Paper, Online, Association for Computational Linguistics, long Paper, to appear. , 2021

Debjit Paul, and Anette Frank "Social Commonsense Reasoning with Multi-Head Knowledge Attention." In Findings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020

Debjit Paul, Juri Opitz, Maria Becker, Jonathan Kobbe, Graeme Hirst and Anette Frank "Argumentative Relation Classification with Background Knowledge." Proceedings of the 8th International Conference on Computational Models of Argument (COMMA 2020) [Nominated Best Paper]

Debjit Paul, and Anette Frank "Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs." Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Paper). 2019.

Debjit Paul, Mittul Singh, Michael A. Hedderich and Dietrich Klakow "Handling Noisy Labels for Robustly Learning from Self-Training Data for Low-Resource Sequence Labeling." Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop. (2019).

Anup Kumar Thander, Goutam Mandal, Debjit Paul "Numerical Comparison of multi-step iterative methods for finding roots of non-linear equations, International Journal of Mathematics Trends and Technology ", Volume4 Issue 8-September 2013 [ISSN: 2231-5373]

About Me

News

Selected Projects

COINS

Generating Hypothetical Events for Abductive Inference

Social Commonsense Reasoning Multi-Head Knowledge Attention

Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs

Handling Noisy Labels for Robustly Learning from Self-Training Data for Low-Resource Sequence Labeling

Argumentative Relation Classification with Background Knowledge

Selected Publications