COMPARISON OF ZERO AND FEW-SHOT LEARNING APPROACH USING THE LLMS FOR SENTIMENT ANALYSIS IN SERBIAN LITERATURE
Domain:
natural language processingRecord type:
paper
Creator:
PetIkoŠkoSta
Publisher:
Host:
Goal: Assess if newer versions of LLMs offer more consistent, efficient, and potentially less biased sentiment annotation.Comparison: With previous research on srpELTeC Sentiment Analysis employing Mistral 7B model.Impact: Advance sentiment data processing and interpretation across various applications.Significance for Low-Resource Languages: Highlight challenges and benefits for languages like Serbian, where annotated corpora are scarce, limiting traditional sentiment analysis models.
Similar
Parameter-Efficient Few-Shot Sentiment Analysis Using LoRA-Enhanced TransformersZero-shot Sentiment Analysis in Low-Resource Languages Using a Multilingual Sentiment LexiconIntermediate-Task Training and Few-Shot Learning for Cross-Domain Robustness in Zero-Shot Cross-Lingual TransferCross-lingual NER Annotation Projection vs. Zero-Shot Few-Shot Learning in Low-Resource Languageskobeoseijnr/Few-Shot-Sentiment-Learning-with-Synthetic-Augmentation-in-Low-Resource-SettingsComparative Study of Different Learning Paradigms for Zero-Shot Sentiment Analysis of the Low-Resource African Language Oromo