Machine Learning and the Library or: How I Learned to Stop Worrying and Love My Robot Overlords

Abstract

Machine learning algorithms and technologies are becoming a regular part of daily life - including life in the libraries. Through this article, I hope to: * To introduce the reader to the basic terminology and concepts of machine learning * To make the reader consider the potential ethical and privacy issues that libraries will face as machine learning permeates society * To demonstrate hypothetical possibilities for applying machine learning to circulation and collections data using TensorFlow/Keras and open datasets Through these goals, it is my hope that this article will inspire a larger, ongoing conversation about the utility and dangers of machine learning in the library (and concurrently society as a whole). In addition, the tripartite division of the article is meant to make the material accessible to readers with different levels of technical proficiency. In approaching the first two goals, the discussion is focused on high level terms and concepts, and it includes specific public cases of machine learning (ab)use that are of broad interest. For the third goal, the discussion becomes more technical and is geared towards those interested in exploring practical machine learning applications in the library.