Hybrid RAG Framework Proposed for Detecting Deceptive Online Reviews
A new arXiv paper presents a retrieval-augmented generation system combining RoBERTa with LoRA fine-tuning, FAISS-based retrieval, and LLM reasoning to identify deceptive online reviews.
Read the full story at arXiv cs.CL ↗