Weinan ZHANG
Ph.D. Student
Department of Computer Science
5th Floor, One Euston Square
Email: w.zhang@cs.ucl.ac.uk
|
I passed my Ph.D. viva in May 2016 and joined Shanghai Jiao Tong University as an assistant professor in Aug. 2016..
Now redirecting to my new homepage in 15 seconds.
Previously I was a Ph.D. student supervised by Dr. Jun Wang and Prof. Stephen Robertson in Media Future Group, Department of Computer Science, UCL. My research interests include machine learning, big data mining and their applications in computational advertising and recommender systems. Particularly, I focus on the research of optimal DSP bidding strategies for RTB display advertising as my Ph.D. direction. I am also interested in deep learning models and have developed several domain-specified DNNs for predicting users' online commercial behaviour. More detailed research descriptions can be found here.
Before coming to UCL, I was a research student in APEX Data and Knowledge Management Lab and a member of ACM Class in Shanghai Jiao Tong University, Shanghai, China. In addition, I was an intern at MediaGamma, Microsoft Research, Google and DERI.
Tutorials
link |
Learning, Prediction and Optimisation in RTB Display Advertising
Weinan Zhang, Jian Xu CIKM, October 2016. |
Real-Time Bidding based Display Advertising: Mechanisms and Algorithms
Jun Wang, Shuai Yuan, Weinan Zhang ECIR, March 2016. |
link |
Paper Collection of Real-Time Bidding
Weinan Zhang Github 2015. |
Research Frontier of Real-Time Bidding based Display Advertising
Weinan Zhang Beijing, August 2015. |
Publications
User Response Learning for Directly Optimizing Campaign Performance in Display Advertising
Kan Ren, Weinan Zhang, Yifei Rong, Haifeng Zhang, Yong Yu, Jun Wang CIKM 2016. (Industry track accept rate: 19.8%) |
LambdaFM: Learning Optimal Ranking with Factorization Machines Using Lambda Surrogates
Fajie Yuan, Guibin Guo, Joemon Jose, Long Chen, Haitao Yu, Weinan Zhang CIKM 2016. (Research track accept rate: 17.6%) |
Functional Bid Landscape Forecasting for Display Advertising
Yuchen Wang, Kan Ren, Weinan Zhang, Jun Wang, Yong Yu ECML/PKDD 2016. |
Bid-aware Gradient Descent for Unbiased Learning with Censored Data in Display Advertising
Weinan Zhang, Tianxiong Zhou, Jun Wang, Jian Xu KDD 2016. (Accept rate: 16.8%) |
Deep Learning over Multi-Field Categorical Data: A Case Study on User Response Prediction
Weinan Zhang, Tianming Du, Jun Wang ECIR 2016. (Accept rate: 21%) |
Implicit Look-alike Modelling in Display Ads: Transfer Collaborative Filtering to CTR Estimation
Weinan Zhang, Lingxi Chen, Jun Wang, Thomas Furmston ECIR 2016. (Accept rate: 21%) |
Collective Noise Contrastive Estimation for Policy Transfer Learning
Weinan Zhang, Ulrich Paquet, Katja Hofmann AAAI 2016. (Accept rate: 25.8%) |
Feedback Control of Real-Time Display Advertising
Weinan Zhang, Yifei Rong, Jun Wang, Tianchi Zhu, Xiaofan Wang WSDM 2016. (Accept rate: 18.2%) |
Risk-Hedged Venture Capital Investment Recommendation
Xiaoxue Zhao*, Weinan Zhang*, Jun Wang RecSys 2015. (Accept rate: 21.4%) |
Statistical Arbitrage Mining for Display Advertising
Weinan Zhang, Jun Wang KDD 2015. (Accept rate: 19.4%) |
Annotating Needles in the Haystack without Looking: Product Information Extraction from Emails
Weinan Zhang, Amr Ahmed, Jie Yang, Vanja Josifovski, Alex J. Smola KDD 2015. (I&G track; accept rate: 34.2%) |
Optimal Real-Time Bidding for Display Advertising
Weinan Zhang, Shuai Yuan, Jun Wang KDD 2014. (Accept rate: 14.6%) |
link |
Real-Time Bidding Benchmarking with iPinYou Dataset
Weinan Zhang, Shuai Yuan, Jun Wang, Xuehua Shen CoRR 2014. abs/1407.7073 |
Bid Keyword Suggestion in Sponsored Search based on Competitiveness and Relevance
Ying Zhang, Weinan Zhang, Bin Gao, Xiaojie Yuan, Tie-Yan Liu IPM. 2014 Vol. 50, No. 4. DOI=10.1016/j.ipm.2014.02.004 |
To Personalize or Not: A Risk Management Perspective
Weinan Zhang, Jun Wang, Bowei Chen, Xiaoxue Zhao RecSys 2013. (Accept rate: 23.8%) |
An Empirical Study of Top-N Recommendation for Venture Finance
Thomas Stone, Weinan Zhang, Xiaoxue Zhao CIKM 2013. (poster) |
Interactive Collaborative Filtering
Xiaoxue Zhao*, Weinan Zhang*, Jun Wang CIKM 2013. (Accept rate: 16.9%) |
Optimizing Top-N Collaborative Filtering via Dynamic Negative Item Sampling
Weinan Zhang, Tianqi Chen, Jun Wang, Yong Yu SIGIR 2013. (short paper) |
MeDetect: A LOD-Based System for Collective Entity Annotation in Biomedicine
Li Tian, Weinan Zhang, Antonis Bikakis, Haofen Wang, Yong Yu, Yuan Ni, Feng Cao WI 2013. |
SVDFeature: A Toolkit for Feature-based Collaborative Filtering
Tianqi Chen, Weinan Zhang, Qiuxia Lu, Kailong Chen, Zhao Zheng, Yong Yu JMLR. 13 (2012) 3619-3622 |
A Semantic Approach to Recommending Text Advertisements for Images
Weinan Zhang, Li Tian, Xinruo Sun, Haofen Wang, Yong Yu RecSys 2012. (Accept rate: 20.2%) |
Joint Optimization of Bid and Budget Allocation in Sponsored Search
Weinan Zhang, Ying Zhang, Bin Gao, Yong Yu, Xiaojie Yuan and Tie-Yan Liu KDD 2012. (Accept rate: 17.6%) |
Serendipitous Personalized Ranking for Top-N Recommendation
Qiuxia Lu, Tianqi Chen, Weinan Zhang, Diyi Yang, Yong Yu WI 2012. Best paper nominee |
MeDetect: Domain Entity Annotation in Biomedical References Using Linked Open Data
Li Tian, Weinan Zhang, Haofen Wang, Chenyang Wu, Yuan Ni, Feng Cao, Yong Yu ISWC 2012. (poster) |
Local Implicit Feedback Mining for Music Recommendation
Diyi Yang, Tianqi Chen, Weinan Zhang, Qiuxia Lu, Yong Yu RecSys 2012. (Accept rate: 20.2%) |
Collaborative filtering with short term preferences mining
Diyi Yang, Tianqi Chen, Weinan Zhang, Yong Yu SIGIR 2012. (poster) |
Advertising Keywords Recommendation for Short-text Web Pages using Wikipedia
Weinan Zhang, Dingquan Wang, Gui-Rong Xue, and Hongyuan Zha ACM TIST. Vol. 3, No. 2. DOI=10.1145/2089094.2089112 |
Feature Based Informative Model for Discriminating Favorite Items from Unrated Ones
Bing Cheng, Tianqi Chen, Diyi Yang, Weinan Zhang, Yongqiang Wang, Yong Yu APWeb 2012. |
TuneSensor: A Semantic-Driven Music Recommendation Service For Digital Photo Albums
Jiansong Chao, Haofen Wang, Wenlei Zhou, Weinan Zhang and Yong Yu ISWC 2011. (poster) |
Informative Household Recommendation with Feature-based Matrix Factorization
Qiuxia Lu, Diyi Yang, Tianqi Chen, Weinan Zhang and Yong Yu RecSys 2011. CAMRa Workshop |
Informative Ensemble of Multi-Resolution Dynamic Factorization Models
Tianqi Chen, Zhao Zheng, Qiuxia Lu, Xiao Jiang, Yuqiang Chen, Weinan Zhang, Kailong Chen, Yong Yu, Nathan N. Liu, Bin Cao, Luheng He and Qiang Yang KDD 2011. KDD-CUP' 11 Workshop |
Feature-Based Matrix Factorization
Tianqi Chen, Zhao Zheng, Qiuxia Lu, Weinan Zhang, Yong Yu CoRR 2011. abs/1109.2271 |
LODDO: Using Linked Open Data Description Overlap to Measure Semantic Relatedness Between Named Entities
Wenlei Zhou, Haofen Wang, Jiansong Chao, Weinan Zhang and Yong Yu JIST 2011. |
Deep Classifier for Large Scale Hierarchical Text Classification
Dingquan Wang,Weinan Zhang, Gui-Rong Xue, and Yong Yu LSHTC 2009. (short paper) |
Internships
MediaGamma Limited Research Intern, Data Science Group, supervised by Jun Wang and Rael Cline. London, United Kingdom Jun. 2015 - Aug. 2015 |
Microsoft Research Cambridge Research Consultant, Machine Learning and Perception Group, supervised by Ulrich Paquet and Katja Hofmann. Cambridge, United Kingdom Feb. 2015 - Jun. 2015 |
Microsoft Research Cambridge Research Intern, Machine Learning and Perception Group, supervised by Ulrich Paquet and Thore Graepel. Cambridge, United Kingdom Sep. 2014 - Nov. 2014 |
Google Inc. Software Engineering Intern, Strategic Technology Group, supervised by Jie Yang and Vanja Josifovski. Mountain View, United States Sep. 2013 - Dec. 2013 |
Digital Enterprise Research Institute Research Intern, Social Software Unit, supervised by Alexandre Passant. Galway, Ireland Jul. 2011 - Sep. 2011 |
Microsoft Research Asia Research Intern, Internet Economics & Computational Ads Group, supervised by Bin Gao and Tie-Yan Liu. Beijing, China Jul. 2010 - Jan. 2011 |
Open Source Projects
OpenBidder An online demand-side platform for benchmarking various RTB algorithms using standard evaluation protocol and real-world ad inventory. Report published in ArXiv. |
SVDFeature An abstract framework to build new matrix factorization variants simply by defining features. Paper published in JMLR-MLOSS. |
Datasets
iPinYou RTB Ad Logs The first public large-scale Real-time bidding advertising log dataset, it is released after iPinYou 2013 RTB competition. We help publish the dataset on the UCL server. Data analysis and benchmarking paper. |
CrunchBase Venture Capital Investments This dataset is collected in May 2014 based on the CrunchBase offical data API. It contains the investment records of VC firms and individual partners. RecSys 2015 paper. |
Competitions
iPinyou DSP 2013 Real-time bidding strategy optimization for demand side platforms. |
CAMERa 2011 Context-aware movie recomendation challenge. |
KDD Cup 2011 Yahoo! music recommendation challenge. |
Yahoo! Learning to Rank Challenge 2010 Ranking top retrieved Web pages from Yahoo! search engine. |
LSHTC 2009 Large-scale deep hierarchy text classification challenge. |
News
9 Aug 2016
Start to work in SJTU. Serveral papers and one tutorial are accepted.
13 May 2016
A full paper "Bid-aware Gradient Descent for Unbiased Learning with Censored Data in Display Advertising" is accepted in KDD 2016.
5 Dec. 2015
Two full papers about user response modelling are accepted in ECIR 2016: 1. "Deep Learning over Multi-Field Categorical Data: A Case Study on User Response Prediction"; 2. "Implicit Look-alike Modelling in Display Ads: Transfer Collaborative Filtering to CTR Estimation".
13 Nov. 2015
My internship work at Microsoft Research Cambridge "Collective Noise Contrastive Estimation for Policy Transfer Learning" is accepted as a full paper in AAAI 2016.
13 Oct. 2015
A full paper "Feedback Control of Real-Time Display Advertising" is accepted in WSDM 2016.
29 Jun. 2015
A full paper "Risk-Hedged Venture Capital Investment Recommendation" is accepted in RecSys 2015.
17 May 2015
My internship work at Google "Annotating Needles in the Haystack without Looking: Product Information Extraction from Emails" is accepted as a full paper in KDD 2015.
13 May 2015
A full paper "Statistical Arbitrage Mining for Display Advertising" is accepted in KDD 2015.
30 Apr. 2015
I will be a program committee member in WSDM 2016.
15 Apr. 2015
We host the 3rd "London Computational Advertising & Behaviour Targeting Group" meetup with the topic "Building Big Data Models and Architecture for RTB Display Advertising". (Link)
23 Feb. 2015
I start to serve as a part-time research consultant on the project of xbox music recommendation at Microsoft Research Cambridge from now to the end of Jun. 2015.
24 Nov. 2014
We host the 2nd "London Computational Advertising & Behaviour Targeting Group" meetup with the topic "Audience Science: User Modelling and Behaviour Targeting". (Link)
25 Jun. 2014
We host the first "London Computational Advertising & Behaviour Targeting Group" meetup with the topic "Programmatic: Promise & Challenges". (Link)
19 Jun. 2014
We give a talk "Optimal Real-Time Bidding for Display Advertising" at Microsoft Research Cambridge.
13 May 2014
A full paper "Optimal Real-Time Bidding for Display Advertising" is accepted in KDD 2014.
20 Apr. 2014
I will be a research intern at Microsoft Research Cambridge from September to November 2014, doing research about large-scale Bayesian inference.
10 Apr. 2014
A journal paper "Bid Keyword Suggestion in Sponsored Search based on Competitiveness and Relevance" is accepted in IPM 2014. This is based on my first internship project at Microsoft.
20 Mar. 2014
We host the data server (located in UK) to publish the iPinYou RTB dataset. (Dataset link) To our knowledge, this is the first published RTB ads research dataset.
31 Dec. 2013
Our team "UCL-CA" wins the champion in "iPinyou Real-Time Bidding DSP Strategy Optimization" contest (season 3). (final ranking)
22 Sep. 2013
Our team "UCL-CA" wins the 3rd position in "iPinyou Real-Time Bidding DSP Strategy Optimization" contest (season 2).
23 Jul. 2013
The work "An Empirical Study of Top-N Recommendation for Venture Finance" is accepted as a 4-page short paper in CIKM 2013. This might be the first work formally bridging collaborative filtering techniques and venture finance.
20 Jul. 2013
A full paper "Interactive Collaborative Filtering" is accepted in CIKM 2013.
10 Jul. 2013
A full paper "To Personalize or Not: A Risk Management Perspective" is accepted in RecSys 2013.
16 Apr. 2013
A 4-page short paper "Optimizing Top-N Collaborative Filtering via Dynamic Negative Item Sampling" is accepted in SIGIR 2013.
28 Feb. 2013
I will be an intern at Google Mountain View from August to November, 2013, doing research about personalization in recommender systems.