1 code implementation • EMNLP 2021 • Viet Lai, Franck Dernoncourt, Thien Huu Nguyen
We address the sampling bias and outlier issues in few-shot learning for event detection, a subtask of information extraction.
no code implementations • Findings (NAACL) 2022 • Luis Guzman-Nateras, Viet Lai, Amir Pouran Ben Veyseh, Franck Dernoncourt, Thien Nguyen
In particular, we introduce SuicideED: a new dataset for the ED task that features seven suicidal event types to comprehensively capture suicide actions and ideation, and general risk and protective factors.
no code implementations • SemEval (NAACL) 2022 • Viet Lai, Amir Pouran Ben Veyseh, Franck Dernoncourt, Thien Nguyen
We describe Symlink, a SemEval shared task of extracting mathematical symbols and their descriptions from LaTeX source of scientific documents.
1 code implementation • Findings (NAACL) 2022 • Viet Lai, Amir Pouran Ben Veyseh, Franck Dernoncourt, Thien Nguyen
This work presents a new human-annotated corpus, called BehancePR, for punctuation restoration in livestreaming video transcripts.
Automatic Speech Recognition Automatic Speech Recognition (ASR) +4
no code implementations • LREC 2022 • Viet Lai, Amir Pouran Ben Veyseh, Franck Dernoncourt, Thien Nguyen
Livestreaming videos have become an effective broadcasting method for both video sharing and educational purposes.
1 code implementation • LREC 2022 • Amir Pouran Ben Veyseh, Viet Lai, Franck Dernoncourt, Thien Nguyen
Question-Answer (QA) is one of the effective methods for storing knowledge which can be used for future retrieval.
no code implementations • 11 Nov 2023 • Rik Koncel-Kedziorski, Michael Krumdick, Viet Lai, Varshini Reddy, Charles Lovering, Chris Tanner
We demonstrate that the current bottleneck in performance is due to LLMs' limited business and financial understanding, highlighting the value of a challenging benchmark for quantitative reasoning within this domain.
no code implementations • ACL 2021 • Amir Pouran Ben Veyseh, Viet Lai, Franck Dernoncourt, Thien Huu Nguyen
To prevent the noises inevitable in automatically generated data from hampering training process, we propose to exploit a teacher-student architecture in which the teacher is supposed to learn anchor knowledge from the original data.