Learning to recommend with data scarcity
How can we learn-to-recommend few-shot items with sparse interactions? This work presents a new perspective on the classic long-tail challenge faced by collaborative filtering models by formulating a meta-learning framework for few-shot recommendation.
"It's all about conversations:" Challenges and Concerns of Faculty and Students in the Arts, Humanities, and the Social Sciences about Education at Scale
Should one teach open-ended courses such as Jazz Improvisation the same way as teaching non-open-ended courses such as Intro to Computer Science? This work presents an exploratory study on the challenges and concerns of faculty and students in the arts, humanities, and social sciences about education at scale.Attribute-guided network sampling mechanisms
How to sample for content (attribute) from Online Social Networks?Transferrable Neural Recommendation and User Preference Modeling
How do we build transferrable deep recommender systems to encode user interests across diverse domains and applications?
Newer