Scaling Recommenders + Content Embeddings at Facebook

This meetup focuses on Scalability and technologies to enable handling large amounts of data: Hadoop, HBase, distributed NoSQL databases, and more!

There's not only a focus on technology, but also everything surrounding it including operations, management, business use cases, and more.

We've had great success in the past, and are growing quickly! Previous guests were from Twitter, LinkedIn, Amazon, Cloudant, Microsoft, 10gen/MongoDB, and more.

This month's guests:

Scaling Recommenders using Apache PredictionIO and the Universal Recommender

This talk covers general scaling problems with recommenders and lays out how they are addressed using a scalable architecture with PredictionIO, Spark, HBase, HDFS, Elasticsearch, and Mahout. Scalability has been addressed at every step from the design of the algorithm to the choice of tools. Several problems are used to illustrate the choices made and issues that others will encounter. In the end are some observations about how Machine Learning should drive application architecture in the future.

BIO: Pat is a committer to Apache PredictionIO and Mahout. He is the Author of the Universal Recommender and Chief Consultant at ActionML. He has helped integrate the UR and other recommenders at scale for several Fortune 500 companies. https://www.linkedin.com/in/pat-ferrel-0051601

Content Embeddings at Facebook

Everyday people create billions of posts, comments, and other kinds of content on Facebook. Understanding that content and making it easily searchable is core to Facebook's mission to make the world more connected. In this talk, we'll talk about how we use learned embeddings for content search and discovery over text, people, and media. The talk will start with a basic introduction to embeddings and cover many novel applications where embeddings are used in production at Facebook.

BIO: Aria Haghighi is an engineering manager at Facebook on content search and discovery. Prior to Facebook, he worked at Apple on [REDACTED], and prior to that was the CTO and Co-Founder of machine learning startup Prismatic. He holds a PhD from UC Berkeley in NLP and ML.

Our format is flexible: We usually have 2 speakers who talk for ~30 minutes each and then do Q+A plus discussion (about 45 minutes each talk) finish by 8:45.

There'll be beer afterwards, of course!

Meetup Location:

Whitepages (http://maps.google.com/maps?q=1301+5th+Avenue+%231700%2C+Seattle%2C+WA), 1301 5th Avenue #1700, Seattle, WA

After-beer Location: Rockbottom is the location for tomorrow’s after-beer c/o Greythorn team.

Doors open 30 minutes ahead of show-time. Please show up at least 15 minutes early out of respect for our first speaker.

Parking is available in the building and is valet only. Cost is $8.00 after 6pm. (Enter on Union between 4th & 5th) Additional parking can be found in the Hilton Parking Garage. Cost is $8.00 after 5pm. Enter on 6th Ave between University and Union. There is also street parking downtown.