ACM RecSys CrowdRec 2015 Workshop

Crowdsourcing and Human Computation for Recommender Systems

ACM RecSys CrowdRec 2015 Workshop

Crowdsourcing and Human Computation for Recommender Systems

Did you mean crowdsourcing for recommender systems?

Omar Alonso, Microsoft

Researchers and practitioners from many areas of computer science have been using crowdsourcing as a quick and inexpensive mechanism to produce labeled data sets. This is usually the case in information retrieval, natural language processing, machine learning, and machine translation to name a few. While the success of recommender systems depends on human computation, unlocking its crowdsourcing potential offers lots of opportunities in terms of research and products. In this talk, I’ll present some lessons learned from both IR and social search and outline the benefits of crowdsourcing in the context of recommender systems.

A community-sourced view of people's surroundings

Henriette Cramer, Yahoo Labs

Physical surroundings are a core element of the human experience. Visual and physical features of places and neighborhoods leave impressions, and can starkly affect behavior and well-being. However, not all aspects of people’s experiences are captured by systems’ local representations. People’s preferences vary, and models based on aggregates of user-generated data can differ from individual perspectives. Using examples from projects at Yahoo Labs such as Urbangems, Henriette will address both opportunities and challenges in using crowd or community-sourced data in creating location-based experiences.

Henriette Cramer is a research scientist in the Mobile Sensing & User Behavior team at Yahoo Labs in California. She focuses on people’s perceptions of personalization and their interactions with mobile, location-based services.