A discussion about why we answer questions, and how a new paradigm can justify crafting search queries that go beyond information seeking.
Photo by Chris Curry on Unsplash
Why do we answer questions? Sometimes it’s to provide information, which has been the interpretation of the computer science community, but in other instances to probe or test intelligence.
In this talk, Dr. Jordan Boyd-Graber, former Associate Professor at the College of Information Studies, reviews the seminal work of Cyril Cleverdon, who conducted a series of breakthrough studies in information retrieval dubbed the Cranfield Experiments, and Alan Turing, the “Father of Computer Science,” who developed the Turing Test. He then presents how their work shaped the information and AI of today and argues that these represent two competing visions for how computers should answer questions: either serving the user (Cranfield) or as an exploration of intelligence (Manchester Paradigm). Boyd-Graber concludes with a discussion of a longer-term plan to achieve the goals of both the Manchester and Cranfield paradigms.
The Manchester Paradigm, proposed by Boyd-Graber and Pedro Rodriguez (Facebook Reality Labs) in their paper, Evaluation Paradigms in Question Answering, aligns closer to the Turing Test, trivia games, and education. The paradigm creates tasks and datasets whose questions push for answers to better understand the world and create evaluations that probe for human-like capabilities. The paper also gives three justifications that highlight three distinct reasons people ask questions beyond information seeking: to teach, to compare, and to probe.
You can watch the full Search Mastery Speaker Series video below or on YouTube here.