Staff Research Scientist at Google
I am aat , based in New York City. I am the research lead and team co-lead for the Responsible ML team, driving research spanning fairness, responsible recommendation and robustness to make Google's products more inclusive and responsible. Previously at Google, I researched recommender systems and reinforcement learning with YouTube and .
I previously got myin computer science at , advised by and . My research focused on large-scale user behavior modeling, covering , , and . Over the course of graduate school, I interned with Facebook's Site Integrity team and News Feed Ranking teams, Microsoft's Cloud and Information Service Lab, and Google Research.
Before graduate school, I majored in computer science and physics at. While there, I worked with and on computational geometry for terrain modeling.
September 2020 - Way out of date! Many new preprints, including new papers on robustness accepted to EMNLP and fairness accepted to NeurIPS.
May 2019 - Our paper on Fairness in Recommendation Ranking through Pairwise Comparisons was accepted to KDD 2019. See you in Anchorage!
January 2019 - I'll be at AIES (and AAAI) to present our works Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements and Counterfactual Fairness in Text Classification through Robustness.
October 2018 - Top-K Off-Policy Correction for a REINFORCE Recommender System, with , , , Francois Belletti, and , was accepted to WSDM 2019.
July 2018 - Categorical-Attributes-Based Item Classification for Recommender Systems, with , , , , Francois Belletti, and , was accepted to RecSys 2018.
May 2018 - Q&R: A Two-Stage Approach Toward Interactive Recommendation, with , , and , was accepted to KDD 2018. See you in London!
March 2018 - Our paper Factorized Recurrent Neural Architectures for Longer Range Dependence was accepted to AISTATS 2018, and I will be presenting the work there in April.
November 2017 - I will be attending the ML Systems workshop at NeurIPS to give a talk on our paper The Case for Learned Indexes.
October 2017 - Our paper Latent Cross: Making Use of Context in Recurrent Recommender Systems was accepted to WSDM 2018.
June 2017 - I will be presenting my recent work, with Jilin Chen, Zhe Zhao, and Ed Chi, at FAT/ML 2017 (at KDD) on fairness properties of adversarial training.
December 2016 - Beyond Globally Optimal: Focused Learning for Improved Recommendations, with Ed Chi, Zhiyuan Cheng, Hubert Pham, and John Anderson, was accepted to WWW 2017.
October 2016 - Recurrent Recommender Networks, with Chao-Yuan Wu, Amr Ahmed, Alex Smola, and How Jing, was accepted to WSDM 2017.
May 2016 - I have completed my Ph.D. (thesis) and will be joining Google Research in Mountain View, CA!
May 2016 - FRAUDAR: Bounding Graph Fraud in the Face of Camouflage, with Bryan Hooi, Hyun Ah Song, Neil Shah, Kijung Shin, Christos Faloutsos, was accepted to KDD 2016 as a full paper with oral presentation.
December 2015 - BIRDNEST: Bayesian Inference for Ratings-Fraud Detection, with Bryan Hooi, Neil Shah, Stephan Gunnemann, Leman Akoglu, Mohit Kumar, Disha Makhija and Christos Faloutsos, was accepted to SDM 2016.
November 2015 - I will be attending NeurIPS in Montreal to co-host the Machine Learning Systems workshop and to present Additive Co-Clustering of Gaussians and Poissons for Joint Modeling of Ratings and Reviews at the workshop on Nonparametric Methods for Large Scale Representation Learning.
September 2015 - I will be speaking at WIN 2015 on CoBaFi - Bayesian collaborative filtering, robust recommendation, and polarized ratings.
August 2015 - A General Suspiciousness Metric for Dense Blocks in Multimodal Data with Meng Jiang, Peng Cui, Christos Faloutsos and Shiqiang Yang was accepted to ICDM 2015.
July 2015 - My tutorial with Leman Akoglu and Christos Faloutsos Fraud Detection through Graph-Based User Behavior Modeling was selected for ACM CCS 2015.
April 2015 - My tutorial with Leman Akoglu and Christos Faloutsos Graph-Based User Behavior Modeling: From Prediction to Fraud Detection was selected for KDD 2015.
April 2015 - I was selected to attend the Heidelberg Laureate Forum in August.
March 2015 - I will be spending the summer at Google Research in Mountain View.
January 2015 - My paper ACCAMS: Additive Co-Clustering to Approximate Matrices Succinctly with Amr Ahmed and Alex Smola was accepted to WWW 2015. I will be in Florence in May to present the work.
November 2014 - My paper Elastic Distributed Bayesian Collaborative Filtering with Markus Weimer, Tom Minka, Yordan Zaykov, and Vijay Narayanan, based on our work this summer at Microsoft, was accepted to the NeurIPS Distributed Machine Learning workshop. I will be at NeurIPS in December to present the work.
October 2014 - My paper Spotting Suspicious Link Behavior with fBox: An Adversarial Perspective with Neil Shah, Brian Gallagher, and Christos Faloutsos has been accepted to ICDM 2014.
July 2014 - CatchSync has been selected as one of the best papers in KDD 2014.
June 2014 - A research proposal that I co-authored with Christos Faloutsos, Amin Mantrach, and Alejandro Jaimes on spam and fraud detection in Tumblr was selected for the Yahoo! Faculty Research and Engagement Program Award.
June 2014 - August 2014 - I am spending the summer at Microsoft, working with the CISL team to scale machine learning on top of REEF.
May 2014 - My paper CatchSync: Catching Synchronized Behavior in Large Directed Graphs with Meng Jiang, Peng Cui, Christos Faloutsos and Shiqiang Yang was accepted to KDD. I will post the camera-ready version soon.
Feb. 2014 - I was lucky enough to win the Facebook Graduate Fellowship for 2014-2015.
Jan. 2014 - The paper Fugue: Slow-Worker-Agnostic Distributed Learning for Big Models with Abhimanu Kumar, Qirong Ho, and Eric Xing was accepted to AISTATS for a full presentation in Reykjavik, Iceland in April.
Jan. 2014 - My paper CoBaFi: Collaborative Bayesian Filtering with Kenton Murray, Alex Smola, and Christos Faloutos was accepted to WWW. I will be presenting it in Seoul, South Korea in April.
Jan. 2014 - The paper FlexiFaCT: Scalable Flexible Factorization of Coupled Tensors on Hadoop with Abhimanu Kumar, Vagelis Papalexakis, Partha Talukdar, Christos Faloutsos, and Eric Xing was accepted to SDM and will be presented in Philadelphia in April. We will relase the source code soon.
Jan. 2014 - The paper Inferring Strange Behavior from Connectivity Pattern in Social Networks with Meng Jiang, Peng Cui, Christos Faloutsos, and Shiqiang Yang was accepted to PAKDD. Meng will be presenting it in Tainan, Taiwan in May.
Top-K Off-Policy Correction for a REINFORCE Recommender System
, Alex Beutel*, , , Francois Belletti,
Latent Cross: Making Use of Context in Recurrent Recommender Systems
Alex Beutel, , Sagar Jain, Can Xu, Jia Li, Vince Gatto,
Recurrent Recommender Networks
, , Alex Beutel, , How Jing
Fairness in Recommendation Ranking through Pairwise Comparisons
Alex Beutel, , Tulsee Doshi, Hai Qian, Li Wei, Yi Wu, Lukasz Heldt, , , , Cristos Goodrow
Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements
Alex Beutel, , Tulsee Doshi, Hai Qian, , Christine Luu, Pierre Kreitmann, Jonathan Bischof,
Counterfactual Fairness in Text Classification through Robustness
, , , , , Alex Beutel
CopyCatch: Stopping Group Attacks by Spotting Lockstep Behavior in Social Networks
Alex Beutel, Wanhong Xu, Venkatesan Guruswami, Christopher Palow, Christos Faloutsos
- ACM Computing Review Editor's Highlight on CopyCatch
- Patent by Facebook (Patent Number 9077744)
- Discussion by Facebook
- Included in courses at Carnegie Mellon and University of Florida
(Best Paper Finalist
in KDD 2014) CatchSync: Catching Synchronized Behavior in Large Directed Graphs
Meng Jiang, Peng Cui, Alex Beutel, Christos Faloutsos, Shiqiang Yang
Code (by Meng) Code + Data
(Best Paper Award)
FRAUDAR: Bounding Graph Fraud in the Face of Camouflage
, , Alex Beutel, , ,
SageDB: A Learned Database System
, , Alex Beutel, , Jialin Ding, Ani Kristo, Guillaume Leclerc, , Hongzi Mao, Vikram Nathan
- Press: The Morning Paper
The Case for Learned Index Structures
, Alex Beutel, , ,
ML Systems at NeurIPS 2017 North East Database Day 2018 SysML 2018
Fugue: Slow-Worker-Agnostic Distributed Learning for Big Models
Abhimanu Kumar, Alex Beutel, Qirong Ho, Eric P. Xing
Appendix Related Presentation