Bridging the gap: A survey of document retrieval techniques for high-resource and low-resource languages

Abstract

With the increasing need for efficient document retrieval in low-resource languages (LRLs), traditional retrieval methods struggle to overcome linguistic challenges such as data scarcity, morphological complexity, and orthographic variations. To address this, hybrid and neural ranking approaches have been explored, integrating statistical retrieval with transformer-based models to enhance search accuracy. Unlike high-resource languages, LRL retrieval requires specialized strategies, including cross-lingual retrieval, domain adaptation, and culturally aware search mechanisms. This article provides a comprehensive review of document retrieval in LRLs, covering classical models, deep learning-based techniques, and their adaptation to resource-constrained languages. A structured taxonomy is introduced, classifying retrieval methods based on model architectures, linguistic processing, and ranking strategies.The paper concludes by highlighting key challenges, benchmarking efforts, and future directions for improving retrieval effectiveness in LRLs.