Comparing Sparse and Dense Information Retrieval Methods on a Wikipedia-Derived NLP Dataset

Abstract

This paper presents a comparative evaluation of sparse and dense Information Retrieval (IR) methods on a domain-focused dataset of 500 NLP-related Wikipedia articles. Four approaches—TF–IDF, BM25, MiniLM, and Dense Passage Retrieval (DPR)—were assessed using Precision@K, nDCG@K, and Hit Rate@K against ground truths from the Wikipedia API and Google results. Results show that BM25 consistently delivers the highest precision and ranking stability across both sources, while TF–IDF remains competitive at larger cut-offs, often surpassing DPR in recall. Dense methods, especially MiniLM, improve recall at higher ranks by capturing semantic relationships, though they lag behind sparse methods in top-rank precision. The divergence between Wikipedia- and Google-based evaluations highlights the importance of multi-perspective benchmarking. Findings confirm the complementary strengths of sparse and dense paradigms, suggesting that hybrid pipelines—combining BM25 with dense re-ranking—provide the most effective balance between efficiency, precision, and semantic coverage for domain-specific retrieval.

More by Iskander Akhmetov

• A Lightweight Extractive Compression Pipeline for Long-Context Retrieval-Augmented GenerationComputación y Sistemas · 2026
• Character N-Gram Features for Kazakh POS Tagging Under Strict OOV EvaluationInternational Conference on Computer Science and Engineering · 2026
• KazakhStopEval: Stop-Word Inventories, Validation Benchmarks, and NLP Evaluation Artifacts for Kazakh Library OCR CollectionsMendeley Data · 2026
• Hierarchical Multi-Agent Extractive Framework for Long-Document Mind Map GenerationComputación y Sistemas · 2026