analiza aiqbench
This paper benchmarks three text data augmentation methods (EDA, back-translation, and contextual substitution) for sentiment classification on small datasets, comparing Logistic Regression, Random Forest, and BERT. The reported pattern is that contextual augmentation helps BERT most consistently, while EDA and back-translation suit traditional classifiers better. For practitioners in low-resource NLP, the useful takeaway is that the best augmentation method depends on the model, so it should be chosen per architecture rather than by default. The results are from relatively targeted experiments, so readers should check dataset coverage and effect sizes before generalizing.
Enhancing Sentiment Classification on Small Datasets through Data Augmentation and Transfer Learning
arXiv:2610.41518v1 Announce Type: new Abstract: Small-scale sentiment classification often suffers from data scarcity, which limits the generalization ability of the models. This study evaluates and compares the effectiveness of three data augmentation strategies: Easy Data Augmentation (EDA), back-translation, and contextual token substitution (nlpaug-style), with both traditional machine learning classifiers (Logistic Regression, Random Forest) and transformer-based models (BERT). We perform a comprehensive empirical comparison with low-resource sentiment datasets by summarizing the results of recent studies and performing targeted head-to-head experiments. Our findings indicate that all augmentation methods improve performance. Contextual augmentation yields the most consistent gains for BERT models, while EDA and back-translation provide greater benefits for traditional classifiers. These insights help guide the selection of data augmentation techniques tailored to model type and dataset size, filling a critical gap in research on data augmentation for sentiment classification on small datasets.
Abstract:Small-scale sentiment classification often suffers from data scarcity, which limits the generalization ability of the models. This study evaluates and compares the effectiveness of three data augmentation strategies: Easy Data Augmentation (EDA), back-translation, and contextual token substitution (nlpaug-style), with both traditional machine learning classifiers (Logistic Regression, Random Forest) and transformer-based models (BERT). We perform a comprehensive empirical comparison with low-resource sentiment datasets by summarizing the results of recent studies and performing targeted head-to-head experiments. Our findings indicate that all augmentation methods improve performance. Contextual augmentation yields the most consistent gains for BERT models, while EDA and back-translation provide greater benefits for traditional classifiers. These insights help guide the selection of data augmentation techniques tailored to model type and dataset size, filling a critical gap in research on data augmentation for sentiment classification on small datasets.
Machine Learning (cs.LG); Computation and Language (cs.CL)
arXiv:2610.41518 [cs.LG]
(or arXiv:2610.41518v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.41518 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
https://doi.org/10.1007/s44163-025-00813-9 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 243 cuvinte
Subiecte conexe
On the Reliability of Generative Augmentation: A Wasserstein-Based Theoretical and Empirical Study
arXiv · 2 sept. 2026
Local data augmentation for imbalanced binary classification
arXiv · 21 oct. 2026
ANN Search: Recall What Matters
arXiv · 5 iun. 2026
Structural Preservation Governs Data Augmentation in Deep Learning-Based Laser Speckle Material Classification
arXiv · 28 iul. 2026