Jiafeng Guo
CAS Key Lab of Web Data Science and Technology
Institute of Computing Technology
Top Conference Review:WSDM 2016
About WSDM
WSDM (pronounced "wisdom") is one of the premier conferences on web inspired
research involving search and data mining.
WSDM publishes original, high quality papers and presentations, with an emphasis on
practical but principled novel models of search, retrieval and data mining, algorithm
design and analysis, economic implications, and in-depth experimental analysis of
accuracy and performance.
About WSDM2016
Conference venue: Mission Bay Conference Center @ San Francisco, USA
Conference Co-Chairs:
Paul N. Bennett, Microsoft Research
Vanja Josifovski, Pinterest
368 submissions, 67 accepted (18.2% acceptance rate).
All accepted papers will appear in the proceedings as full-length publications and
will also be presented as both a talk, and a poster during an interactive session.
Overview
Schedule:
3 keynotes
4 invited talks for Practice & Experience Track
paper presentations
5 workshops
3 tutorials
a doctoral consortium
a VC/Industry Day
WSDM CUP
To provide the best static rank values for each of publication entity in a
heterogeneous graph (Microsoft Academic Graph)
Papers
Best Paper:
Beyond Ranking: Optimizing Whole-Page Presentation
Yue Wang, Dawei Yin, Roger Jie Luo, Pengyuan Wang, Makoto Yamada, Yi Chang
and Qiaozhu Mei
Honorable Mentions
Information Evolution in Social Networks
Lada A. Adamic, Thomas M. Lento, Eytan Adar and Pauline C. Ng
DiFacto: Distributed Factorization Machines
Mu Li, Ziqi Liu, Alexander Smola and Yu-Xiang Wang
Keynote @ Industry Day
Keynote by Ron Conway (Founder & Co-managing Partner, SV Angel)
Driving Innovation with Data Science
“A truly great founder is product focused & determined”
Keynote by Qi Lu (Executive Vice President, Microsoft)
The Anatomy of an Emerging Digital Society: A Look into the
Future from an Industrial Development Perspective
The 1st Era: PC and Client/Server
The 2nd Era: Internet and Web
The 3rd Era: Mobile and Cloud
“Conversation is the platform.” Natural language is the
ultimate UI.
Panels @ Industry Day
Panel 1: The role of data science in startups
Panel 2: How to transform business with data science
Panel 3: Data Science and Machine Learning in the Sharing Economy
Keynote @ First Day
Keynote by Jeff Dean (Google Senior Fellow, Google Research)
Large-Scale Deep Learning For Building Intelligent
Computer Systems
Google Brain
TensorFlow
Invited Talk @ First Day
Talk by Yoelle Maarek (Vice-President of Research, Yahoo)
Is Mail The Next Frontier In Search And Data Mining?
Email ranking by considering “actions”
Automated labels/Left categories across many users
Predicting future emails based on sequences
Keynote @ Second Day
Keynote by Yiling Chen (Gordon McKay Professor of Computer
Science, Harvard University)
Why Incentive Alignment is Relevant for Data Science
“what data become available depends on the interaction rules”
“jointly considers data acquisition and inference and learning
is important for data science and machine learning”
Invited Talks @ Second Day
Talk by Jie Tang (Associate Professor, Tsinghua University)
AMiner: Toward Understanding Big Scholar Data
Academic data mining
Talk by Lars Backstrom (Director of Engineering on News
Feed Facebook)
Serving a Billion Personalized News Feeds
Personalized feed ranking
Trending 1
It is critical to understand Users
Understanding User Attention and Engagement in Online News Reading (Social
Media)
Modeling and Predicting Learning Behavior in MOOCs (MOOC)
Your Cart tells You: Inferring Demographic Attributes from Purchase Data (Retail)
You've got Mail, and Here is What you Could do With It!: Analyzing and
Predicting Actions on Email Messages (Email)
Many papers discuss about how to model user behaviors, predict users’actions,
capture users’ preferences, enhance user engagement, and provide
personalized ranking/recommendations.
Trending 2
From Small to Big
Scaling up Link Prediction with Ensembles
DiFacto: Distributed Factorization Machines
Distributed Balanced Partitioning via Linear Embedding
Multi-view Machines
Multileave Gradient Descent for Fast Online Learning to Rank
A focused session on Big Data Algorithms, talking about large scale distributed
models and online learning algorithms.
Some feeling & thoughts
Very diverse topics in accepted papers
Mobile, search , recommendation, crowdsourcing,
privacy…
Deep learning is not dominant, but representation learning is
arising
Social computing is very popular
Social media, social network, communities, MOOC…
Mining, search, quality control, behavior prediction…
Thanks!