CS 201: Big Data and Data Science in Taboola’s Newsroom Product, MATT (YIXIONG) MENG – JUN XU, Taboola

Speaker: Matt [Yixiong] Meng - Jun Xu
Affiliation: Taboola

ABSTRACT: Taboola is one of fastest-growing global tech companies. We empower hundreds of millions of people and thousands of publishers to personalize the Web. We enable publishers to drive editorial voice and brand safety, while giving consumers the choice to personalize their content discovery experience. We engage over 1B online users, and deliver more than 300B content recommendations every month! We are a based in Israel, with offices globally in NYC, Los Angeles, Bangkok, New Delhi, Sao Paulo, and more. This lecture will specifically cover Taboola’s Newsroom product, with a focus on the data science-centric Network Insights module. We will discuss the overall Big Data pipeline and data science used in the Taboola Newsroom product . The pipeline discussion will focus on Spark and Spark Streaming, and the data science discussion will focus on topic, discovery, using document embedding and clustering technologies. A product demo will be provided, followed by Q&A. BIO: Matt (Yixiong) Meng is currently a Data Scientist, Machine Learning at Taboola. He holds a PhD from the Physics and Astronomy Department of University of California Los Angeles. He is passionate in providing solutions with machine learning models. At Taboola, he applies Recurrent Neural Network (RNN) in time series prediction, and also builds models using seq2seq framework in text comprehension. Previously he worked as a data scientist at Avant and built machine learning models to make instant underwriting decisions. BIO: Jun Xu was a seasoned software engineer with 15 years experience in Java development. Five years ago he switched his interest and passion to data science. His main specialty currently is in big data, predictive modeling and text data mining. He has been a Principal Engineer at Taboola for over three years and currently is a principal member of Taboola’s data team in LA. He currently works on both Python and Java stack using technologies like Spark, Google Bigquery, TensorFlow and Google ML-Engine etc. Prior to his time at Taboola, he has worked for various other companies such as Perfect Market (as Principal System Architect) and Sun Microsystems, which is now Oracle (as Senior Software Engineer). 

Hosted by Professor Paul Eggert

 

 

Date/Time:
Date(s) - Nov 28, 2017
4:15 pm - 5:45 pm

Location:
3400 Boelter Hall
420 Westwood Plaza Los Angeles California 90095