How to build a recommendation engine, for wine.
===== TALK OUTLINE =====
Wine pairing has long been an art and a science, dependent upon an elite palate and firm grasp of experience. But it doesn't have to be that way. In this talk, Joseph Nelson discusses how data science enables individuals, even those with little subject area expertise, to produce functioning recommendation engines. Upon an award from the French Embassy, he led a team that sought to use data science to build a functioning wine recommendation engine. He describes clustering, user segmentation, and modeling to provide accurate pairings. While the model is not currently in deployment, Joseph discusses how French culinary experts reacted to an unveiling of the demonstration.
====== SPEAKER ======
Joseph is currently an instructor for General Assembly’s Data Science Immersive in DC. He particularly enjoys model selection and validation techniques as well as natural language processing. He has also guest lectured as a part of the Master’s in Data Science program at George Washington University. Previously, he built data visualization dashboards and data pipelines as a part of Facebook’s Politics and Government team. Joseph is an advocate for using technology to improve social well-being, particularly in government. When he’s not modeling (data), he’s reading, running, or learning the tools to build his next side project.
You can find him on Twitter: https://twitter.com/josephofiowa and Github: github.com/josephofiowa
-- TENTATIVE SCHEDULE --
6:30-7:00 - Networking
7:00-7:10 - Introductions
7:10 - 8:10 - Talk
> 8:10- Socializing