| 2023 |
The Best of Both Worlds: Combining Human and Machine Translations for Multilingual Semantic Parsing with Active Learning |
Li et al. |
ACL |
- |
Multilingual Semantic Parsing |
Hybrid, BERT-LSTM, None, PT+FT, Hard |
GEOQUERY, NLMAP |
Most semantically diversified and representative utterances improves the parser performance to the greatest extent. |
| 2023 |
Counterfactual Active Learning for Out-of-Distribution Generalization |
Deng et al. |
ACL |
- |
Dissonance Detection |
Uncertainty, RoBERTa-based, Transfer Learning, PT+FT, Hard |
build a dissonance dataset |
|
| 2023 |
Transfer and Active Learning for Dissonance Detection: Addressing the Rare-Class Challenge |
Vasudha Varadarajan et al. |
ACL |
- |
Sentiment Analysis and Natural language Inference |
Informative, RoBERTa, Counterfactual thinking, PT+FT, Hard |
IMDB, |
|
| 2023 |
Deep Active Learning for Morphophonological Processing |
Seyed Morteza Mirbostani et al. |
ACL |
- |
morphological processing |
informative, DNNs, None, PT+FT, Hard |
|
Arabic morphophonology dataset |
| 2023 |
On Dataset Transferability in Active Learning for Transformers |
Fran Jelenić et al. |
ACL |
- |
text classification |
Uncertainty, PLMs, None, PT+FT, Hard |
Subjectivity, CoLA, AG-NEWs, and TREC |
We investigate how ASM affects dataset transferability and how ASM is affected by other AL variables. |
| 2023 |
On the Limitations of Simulating Active Learning |
Katerina Margatina et al. |
ACL |
- |
|
Any, LLMs, None, PT+FT, Hard |
|
Simulating,why do ac- tive learning algorithms sometimes fail to out- perform random sampling |
| 2023 |
D-CALM: A Dynamic Clustering-based Active Learning Approach for Mitigating Bias |
Sabit Hassan et al. |
ACL |
- |
text classification tasks |
clustering, PLMs, None, PT+FT, Hard |
BOOK32, CONAN, CARER, CoLA, HATE, MRDA, Q-Type, Subjectivity |
infusing clustering with AL can overcome the bias issue of both AL and traditional annotation methods while exploit- ing AL’s annotation efficiency. |
| 2023 |
Which Examples Should be Multiply Annotated? Active Learning When Annotators May Disagree |
Connor Baumler et al. |
ACL |
- |
Text Classification |
disagreement, PLMs, None, PT+FT, Hard |
Measuring Hate Speech (MHS) (Sachdeva et al., 2022) and Wikipedia Talk |
|
| 2023 |
Reinforced Active Learning for Low-Resource, Domain-Specific, Multi-Label Text Classification |
Lukas Wertz et al. |
ACL |
- |
Text classification |
Any, BERT, Reinforcement,PT+FT, Hard |
ArXiv, EurLex57k, Patents, Yelp, |
Reinforced Active Learning for Low-Resource, Domain-Specific, Multi-Label Text Classification |