Artificial intelligence (AI) and machine learning technology are transforming a whole host of industries, from healthcare to marketing and finance—and they have the potential to do the same for academic libraries.

Machine learning is a subset of AI that gives computer systems the ability to learn and improve from prior experience automatically, without being explicitly programmed. The technology’s power “lies in the fact that machines can recognize patterns efficiently and routinely, at a scale and speed that humans cannot approach,” writes Catherine Nicole Coleman, digital research architect for Stanford University.

In the world of finance, AI and machine learning are used to analyze a large number of disparate data sets to identify risk and help investors make smart decisions. In health care, the technology is used to design the best treatment plans and to recommend personalized care based on patient information. Businesses use AI to analyze massive amounts of data on customer preferences and behaviors to know which products they should market to whom.

So far, AI’s potential has remained largely untapped among research libraries. A recent Ex Libris survey revealed that while nearly 80 percent of research librarians are exploring the use of AI and machine learning, only about 5 percent are currently leveraging the technology.

A number of concerns seem to be holding back the use of AI among libraries. Aside from the usual budgetary pressures, there is some fear the technology might make librarians’ jobs obsolete or help propagate misinformation if there are problems with the algorithms that power it.

While these are legitimate concerns, AI can help academic libraries demonstrate real value to their institution if it is used judiciously. For instance, the technology can help produce more individually tailored (and personally relevant) search results and resource recommendations for students and researchers. It can also improve the operational efficiency of library staff, thereby freeing up librarians to focus on higher-level tasks.

These enhancements are important, because they come at a time when academic libraries are struggling to define their roles in the digital age.

With powerful search engines just a click away, today’s students (and even researchers) view themselves as increasingly self-sufficient in their information gathering skills. In a world in which Google Scholar and other online resources are available free of charge, libraries are looking for new ways to establish value for their stakeholders.

One practical example of how AI can help support the work of academic libraries is an intelligent decision-support tool embedded within the Ex Libris Alma cloud-based library services platform.

This data analysis and recommendation feature helps libraries become more efficient by providing staff with recommendations to improve their workflows and better optimize their use of the Alma platform.

For example, the system periodically reviews the Counting Online Usage of Networked Electronic Resources (COUNTER) files that librarians have manually uploaded to Alma to keep track of how resources are being used, and it tries to determine if this information can be collected in an automated fashion instead.

If the system recognizes that other institutions have successfully harvested usage data from a particular content vendor using the Standardized Usage Statistics Harvesting Initiative (SUSHI) protocol, it will suggest that librarians create a SUSHI account for that vendor to harvest data on the use of that vendor’s resources electronically.

By automating this workflow, the librarian can reduce their chances of making a mistake—while saving valuable time that can be spent on more strategic priorities instead, such as providing a better user experience, attracting new and more diverse audiences, and helping students develop their research skills.

Rather than viewing AI as a destructive force that will replace their jobs or disrupt traditional processes, librarians should think of it as a tool to solve key challenges. With the help of intelligent technologies, academic libraries can improve efficiencies, save costs, show a measurable return on investment, and serve a new generation of users more effectively. In short, they can reinvent themselves for a new era while providing additional value for stakeholders.

For more information about AI’s potential in supporting research libraries, see the white paper: “Artificial Intelligence in the Library: Advantages, Challenges, and Tradition.”