A Simple and Effective Model for Answering Multi-span Questions

Segal, Elad; Efrat, Avia; Shoham, Mor; Globerson, Amir; Berant, Jonathan

doi:10.48550/arxiv.1909.13375

Cited by 3 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the question answering task, the question does not require labels. (Segal et al, 2019) also considers the multi-answer question problem as a sequence tagging problem, and only the answer is tagged with labels. Huggingface 10 mentions, if a token with label -100, it will not be considered in entropy loss calculation.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

GUDN: A novel guide network with label reinforcement strategy for extreme multi-label text classification

Wang

Zhu

Shu

et al. 2023

Journal of King Saud University - Computer and Information Scie

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

“…Huggingface 10 mentions, if a token with label -100, it will not be considered in entropy loss calculation. Therefore, the experiment uses -100 to label sentence 1 and uses IO tagging, which performs better than BIO in multi-answer questioning (Segal et al, 2019) to label sentence 2.…”

Section: Related Workmentioning

confidence: 99%

GUDN: A novel guide network with label reinforcement strategy for extreme multi-label text classification

Wang

Zhu

Shu

et al. 2023

Journal of King Saud University - Computer and Information Scie

View full text Add to dashboard Cite

“…For extractive reading comprehension such as SQuAD (Rajpurkar et al, 2016), the answer is a span in the text, and the MRC model ) gets the answer by predicting the probability that the word is start or end. Some datasets such as DROP (Dua et al, 2019) have answers that include multiple spans, and the answers can be obtained by using BIO tagging (Segal et al, 2019). For multiple-choice reading comprehension where the answer is one of several options, a method (Pan et al, 2019) is to calculate the score for each option and then select the option with the highest score.…”

Section: Machine Reading Comprehensionmentioning

confidence: 99%

An MRC Framework for Semantic Role Labeling

Wang¹,

Li²,

Meng³

et al. 2021

Preprint

View full text Add to dashboard Cite

Semantic Role Labeling (SRL) aims at recognizing the predicate-argument structure of a sentence and can be decomposed into two subtasks: predicate disambiguation and argument labeling. Prior work deals with these two tasks independently, which ignores the semantic connection between the two tasks. In this paper, we propose to use the machine reading comprehension (MRC) framework to bridge this gap. We formalize predicate disambiguation as multiple-choice machine reading comprehension, where the descriptions of candidate senses of a given predicate are used as options to select the correct sense. The chosen predicate sense is then used to determine the semantic roles for that predicate, and these semantic roles are used to construct the query for another MRC model for argument labeling. In this way, we are able to leverage both the predicate semantics and the semantic role semantics for argument labeling. We also propose to select a subset of all the possible semantic roles for computational efficiency. Experiments show that the proposed framework achieves state-of-the-art results on both span and dependency benchmarks. 1

show abstract

“…Transformer based architectures have also produced state-of-the-art performance on sequence tagging tasks like Named Entity Recognition (NER) (Yamada et al, 2020;Devlin et al, 2019; span extraction (Eberts and Ulges, 2019;Joshi et al, 2020) and QA tasks (Devlin et al, 2019;Lan et al, 2020). Multiple span extraction from texts has been explored both as a sequence tagging task (Patil et al, 2020;Segal et al, 2019) and as span extraction as in RC tasks (Hu et al, 2019;.…”

Section: Literaturementioning

confidence: 99%

Cisco at SemEval-2021 Task 5: What’s Toxic?: Leveraging Transformers for Multiple Toxic Span Extraction from Online Comments

Ghosh¹,

Kumar²

2021

Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021)

View full text Add to dashboard Cite

Social network platforms are generally used to share positive, constructive, and insightful content. However, in recent times, people often get exposed to objectionable content like threat, identity attacks, hate speech, insults, obscene texts, offensive remarks or bullying. Existing work on toxic speech detection focuses on binary classification or on differentiating toxic speech among a small set of categories. This paper describes the system proposed by team Cisco for SemEval-2021 Task 5: Toxic Spans Detection, the first shared task focusing on detecting the spans in the text that attribute to its toxicity, in English language. We approach this problem primarily in two ways: a sequence tagging approach and a dependency parsing approach. In our sequence tagging approach we tag each token in a sentence under a particular tagging scheme. Our best performing architecture in this approach also proved to be our best performing architecture overall with an F 1 score of 0.6922, thereby placing us 7 th on the final evaluation phase leaderboard. We also explore a dependency parsing approach where we extract spans from the input sentence under the supervision of target span boundaries and rank our spans using a biaffine model. Finally, we also provide a detailed analysis of our results and model performance in our paper.

show abstract

A Simple and Effective Model for Answering Multi-span Questions

Cited by 3 publications

References 0 publications

GUDN: A novel guide network with label reinforcement strategy for extreme multi-label text classification

GUDN: A novel guide network with label reinforcement strategy for extreme multi-label text classification

An MRC Framework for Semantic Role Labeling

Cisco at SemEval-2021 Task 5: What’s Toxic?: Leveraging Transformers for Multiple Toxic Span Extraction from Online Comments

Contact Info

Product

Resources

About