Corpus ID: 232076220

BERT-based Acronym Disambiguation with Multiple Training Strategies

@article{Pan2021BERTbasedAD,
  title={BERT-based Acronym Disambiguation with Multiple Training Strategies},
  author={Chunguang Pan and Bingyan Song and S. Wang and Zhipeng Luo},
  journal={ArXiv},
  year={2021},
  volume={abs/2103.00488}
}
Acronym disambiguation (AD) task aims to find the correct expansions of an ambiguous ancronym in a given sentence. Although it is convenient to use acronyms, sometimes they could be difficult to understand. Identifying the appropriate expansions of an acronym is a practical task in natural language processing. Since few works have been done for AD in scientific field, we propose a binary classification model incorporating BERT and several training strategies including dynamic negative sample… Expand

Figures and Tables from this paper

Acronym Identification and Disambiguation Shared Tasks for Scientific Document Understanding
TLDR
Two shared task for acronym identification and acronym disambiguation in scientific documents, named AI@ SDU and AD@SDU, respectively are organized and the prominent participating systems for each of them are reviewed. Expand

References

SHOWING 1-10 OF 32 REFERENCES
A Supervised Learning Approach to Acronym Identification
TLDR
This paper reduces the search space of the supervised learning system by putting some weak constraints on the kinds of acronym-definition pairs that can be identified and obtains results comparable to hand-crafted systems that use stronger constraints. Expand
Using Word Embeddings for Unsupervised Acronym Disambiguation
TLDR
This work presents a method to choose the contextual correct definition of an acronym that does not require training for each acronym and thus can be applied to a large number of different acronyms with only few instances. Expand
Acronyms: identification, expansion and disambiguation
TLDR
This work presents a new machine-learning-based approach to automatically build an acronym dictionary from unannotated texts, and is the first such technique that specifically handles non-local acronyms, i.e., that can determine an acronym’s expansion even when the expansion does not appear in the same document as the acronym. Expand
What Does This Acronym Mean? Introducing a New Dataset for Acronym Identification and Disambiguation
TLDR
A new deep learning model is proposed which utilizes the syntactical structure of the sentence to expand an ambiguous acronym in a sentence and outperforms the state-of-the-art models on the new AD dataset, providing a strong baseline for future research on this dataset. Expand
Guess Me if You Can: Acronym Disambiguation for Enterprises
TLDR
This work proposes an end-to-end framework to automatically resolve the true meanings of acronyms in a given context and produces a high-quality acronym disambiguation system as output. Expand
Acronym Identification and Disambiguation Shared Tasks for Scientific Document Understanding
TLDR
Two shared task for acronym identification and acronym disambiguation in scientific documents, named AI@ SDU and AD@SDU, respectively are organized and the prominent participating systems for each of them are reviewed. Expand
Multi-granularity sequence labeling model for acronym expansion identification
TLDR
A Latent-state Neural Conditional Random Fields model (LNCRF) is proposed to deal with the challenges of finding expansions in texts for given acronym queries by modeling the problem as a sequence labeling task. Expand
Abbreviation Explorer - an interactive system for pre-evaluation of Unsupervised Abbreviation Disambiguation
TLDR
Abbreviation Explorer is a system that supports interactive exploration of abbreviations that are challenging for Unsupervised Abbreviation Disambiguation (UAD) and provides options for creating corrective rules that merge redundant long-forms with identical meaning. Expand
Part-of-speech tagging with antagonistic adversaries
TLDR
A new perceptron learning algorithm using antagonistic adversaries is presented and compared to previous proposals on 12 multilingual cross-domain part-of-speech tagging datasets, with an average 4% error reduction. Expand
Deep Unordered Composition Rivals Syntactic Methods for Text Classification
TLDR
This work presents a simple deep neural network that competes with and, in some cases, outperforms such models on sentiment analysis and factoid question answering tasks while taking only a fraction of the training time. Expand
...
1
2
3
4
...