New AI software like ChatGPT challenges traditional information retrieval models and creates new opportunities. The ability to understand natural language and develop interactions can make information retrieval with ChatGPT more efficient and intuitive.
The potential of AI in information retrieval can include:
- More ideas for search terms à Replaces the use of reference books.
- Condenses broad topics à Need for source criticism in academic information retrieval increases.
- Helps to find the main point of view à Brings out new perspectives and ideas.
However, there are still some challenges when using AI in information retrieval:
- ChatGPT is not yet able to perform searches in scholarly databases.
- Condenses broad topics à Need for source criticism in academic information retrieval increases.
- ChatGPT does not cite sources à Possibility of plagiarism increases.
- ChatGPT has been found to generate studies that do not exist.
- So far, ChatGPT data is not automatically updated and it has limited knowledge of events that occurred after September 2021.
Links:
Is ChatGPT making up references?
ChatGPT in Healthcare: A Taxonomy and Systematic Review
University of Calgary - library guide on AI
