Search

The introduction of ChatGPT 3.5 in November 2022 ignited a sensation in the academic community, leaving many astounded by its capabilities. This new release more closely emulates human responses than its predecessors. Among its remarkable capabilities, it can answer questions, catalog items in MARC21, recommend reading lists, and make suggestions on a wide array of topics. To assess ChatGPT’s efficacy in aiding library users, the authors of this paper conducted an experiment comparing ChatGPT’s performance with that of librarians in answering reference questions. Thirty questions were randomly selected from the transaction log of the reference inquiries between June 1, 2023 to July 31, 2023 at the Rider University Libraries. These queries constituted 34% of the total user questions during this two-month period. The authors compared the answers by ChatGPT and those by reference librarians for their accuracy, relevance, and friendliness. The findings indicate that reference librarians markedly outperformed their robotic counterpart. An evident issue arises from ChatGPT’s deficiency in understanding local policies and practices. This consequently hinders its ability to provide satisfactory answers in those areas. OpenAI posits that ChatGPT’s proficiency can be enhanced through targeted fine-tuning using locally specific information. At the moment, ChatGPT remains a great tool for librarians. [ABSTRACT FROM AUTHOR]

The research aimed to explore the perspectives of students toward the utilization of ChatGPT, its potential benefits and disadvantages over traditional reference librarians. The study employed a descriptive survey research methodology and analyzed the data collected through descriptive and inferential statistics. Out of the 54 students at Adeleke University surveyed, they acknowledged the advantages of ChatGPT such as its convenience, user-friendliness, extensive knowledge base, and accessibility. However, they also pointed out its limitations including the inability to comprehend emotions, limitations in answering complex questions, potential for providing incorrect answers, and risk of outdated information. Despite these drawbacks, the students still expressed their intention to seek reference assistance from ChatGPT, as well as from librarians. The research suggests that library administration should embrace technology to ensure that students have convenient access to reference services. [ABSTRACT FROM AUTHOR]

This article explores Internet search tools that harness emerging technologies and user-centric features such as Web 2.0 sites, mobile applications, metasearching, real-time information, peer-to-peer searches, visual interfaces, and computational abilities. Reference librarians everywhere have incorporated Google into their searching repertoires, but what about other search engines that feature 'bleeding edge' technologies and human-friendly designs? Our article is a concise guide to nine innovative search engines-Silobreaker, Zuula, Bing, ChaCha, CompletePlanet, DuckDuckGo, Spezify, Wolfram|Alpha, and Wowd-which give us a glimpse into the future of searching technology. We conclude our article by discussing implications for reference services. [ABSTRACT FROM AUTHOR]