by Carla Pfahl
Quick Summary
In an effort to ensure any AI tools available through our hosted e-resources align with our values, Minitex has used the REACT Framework to evaluate all AI features in our library databases.
Developed by Loyola Marymount University (LMU) Library, the REACT Framework is a model for evaluating generative AI tools in academic research settings. It uses a 1-4 rating scale to assess various aspects of generative tools, with 1 being the lowest and 4 the highest. The purpose is to evaluate AI tools so researchers can make information decisions and adopt tools that best meet their needs. Tools are assessed by:
- Relevance - How well the tool aligns with your research goals and addresses your needs.
- Ease of Use - How intuitive and user-friendly the tool is.
- Assessing DEIA - Whether the tool is inclusive, unbiased, and accessible to all users.
- Currency - Whether the tool uses up-to-date data and current AI best practices.
- Transparency & Accuracy - Whether the tool is clear about how it works and consistently delivers reliable information.
Currently, two eLibraryMN (ELM) vendors have added generative AI tools to their ELM resources as either opt-in or by default option. Minitex Outreach & Instruction librarians used the REACT Framework to highlight AI tools available through eLibraryMN to provide a better understanding of their function.
We have reviewed the resources and published our findings and scores through the Educators & Librarians section of the ELM website. A new page, Assessing Library AI Features with the REACT Framework, has been created to make that information available to the public.
