Prototyping expert systems in reference services (1980–2000): experimentation, success, disillusionment, and legacy.

In the late twentieth century librarians prototyped expert systems in reference services in order to respond to the reference 'crisis' of the time and to harness the power of emerging artificial intelligence (AI) technologies. Creating intelligent systems required librarian designers to codify the expertise of reference librarians and the resources of reference services into knowledge representation mechanisms suitable for inferences by the system. In this process, librarians explored the theoretical and pragmatic bases of reference and experimented with how to implement them in AI. The successes and failures of these prototypes reveal how librarians felt about these new technologies and how they might transform libraries. While expert systems in libraries failed and were abandoned by the end of the century, lessons and insights from this seminal work can inform current activities in the application of machine learning. [ABSTRACT FROM AUTHOR]