Speech-Driven Information Retrieval Systems: A Review of AI and NLP Techniques

Abstract

This paper offers a critical review of the progress, issues, and future of speech-driven information retrieval (SDIR). Following ten studies as references, this paper focuses on the use of artificial intelligence (AI) and natural language processing (NLP) in support of speech-based interaction within information retrieval systems (IRSs). Innovative difficulties, including differences in the accents, noise, and the peculiarity of the words, and the contextual and multilingual approach are also mentioned. This paper discusses how current trends in AI, such as transformers like GPT and BERT, restore in-depth features for the enhancement of the speech recognition rate, semantic content analysis, and operational queries in real time. Furthermore, future trends, including multilingual retrieval systems and real-time processing, are examined as significant advancements in improving the SDIR systems’ accessibility and speed. Overcoming these challenges and building on advances in AI, the study aims towards the development of future SDIR systems that offer optimal, easy, and versatile solutions for various uses