Keyword Searching and Digital Archives on Web: Challenges and Innovations in GLAM

Introduction; In the evolving digital landscape, keyword searching plays a pivotal role in facilitating access to information stored in cultural heritage archives. Objective; This paper explores the current challenges and recent innovations in keyword search technologies within the Galleries, Libraries, Archives, and Museums (GLAM) sector, emphasizing web-based retrieval systems. With the growth of digital archives such as Europeana and the Digital Public Library of America (DPLA), institutions face complexities in semantic search, multilingual access, and metadata standardization. Method; We evaluate traditional keyword models like TF-IDF against advanced AI-based approaches such as BERT, focusing on their effectiveness in web contexts. Result; Through case studies and performance evaluations, we identify promising methodologies that improve semantic relevance and user accessibility. Conclusion; The findings reveal that BERT-based models significantly outperform legacy methods, particularly in multilingual and semantically ambiguous search environments. The paper concludes with strategic recommendations for implementing AI-driven keyword search frameworks in GLAM archives.