Evaluation of Chunking Strategies for Effective Text Embedding in Low-Resource Language on Agricultural Documents

In three linesComparative study of four chunking strategies (Recursive, Khmer-Aware, Sentence-Based, LLM-Based) for RAG on Khmer agricultural documents. Recursive chunking with 300 characters achieves best performance: L2 distance 0.4295, Answer Relevance 0.8663, Khmer IoU 0.6441. Statistically significant improvement over Sentence-Based (p=0.0121).

Summary generated by Claude — human-verified