AI-Powered academic search systems: An exploratory study from the perspective of academics in LIS

Abstract

The study explored the reality and benefits of the use of AI-powered academic search systems from the perspective of Egyptian academics in Library and Information Science (LIS). It also provided a comprehensive view of this category of systems. The study used a self-administered questionnaire as a survey instrument to gather data from the study population, which consisted of faculty staff and assistants belonging to the LIS departments (22 departments) in Egypt. The sample included 164 participants who responded to the questionnaire. SPSS was used to perform the statistical analyses. The results showed that 42.7% of participants used AI-powered academic search systems, while the majority (57.3%) did not. The most prominent reason for not using AI-powered academic search systems was the preference of most participants for alternative search systems, such as library catalogs, online databases, and search engines. Overall, Semantic Scholar (41.43%), Scite (37.14%), and Research Rabbit (32.86%) were the most commonly used systems. Moreover, the study demonstrated a positive level of participant satisfaction with these systems, with means from 3.614 to 4.257 and standard deviations from 0.700 to 1.067. It revealed the most significant concerns that users face when interacting with this category of systems. Finally, these concerns were discussed in detail and some solutions were presented to address them.