Google AI Blog: August 2015

VLDB 2015 and Database Research at Google

Monday, August 31, 2015

Posted by Corinna Cortes, Head of Google Research NY and Cong Yu, Research Scientist41st International Conference of Very Large Databasesstructured snippetstable searchF1MesaBigQuerybluePapers:Keys for GraphsWenfei Fan, Zhe Fan, Chao Tian, Xin Luna DongIn-Memory Performance for Big DataGoetz Graefe, Haris Volos, Hideaki Kimura, Harumi Kuno, Joseph Tucek, Mark Lillibridge, Alistair VeitchThe Dataflow Model: A Practical Approach to Balancing Correctness, Latency, and Cost in Massive-Scale, Unbounded, Out-of-Order Data ProcessingTyler Akidau, Robert Bradshaw, Craig Chambers, Slava Chernyak, Rafael Fernández-Moctezuma, Reuven Lax, Sam McVeety, Daniel Mills, Frances Perry, Eric Schmidt, Sam WhittleResource Bricolage for Parallel Database SystemsJiexing Li, Jeffrey Naughton, Rimma NehmeAsterixDB: A Scalable, Open Source BDMSSattam Alsubaiee, Yasser Altowim, Hotham Altwaijry, Alex Behm, Vinayak Borkar, Yingyi Bu, Michael Carey, Inci Cetindil, Madhusudan Cheelangi, Khurram Faraaz, Eugenia Gabrielova, Raman Grover, Zachary Heilbron, Young-Seok Kim, Chen Li, Guangqiang Li, Ji Mahn Ok, Nicola Onose, Pouria Pirzadeh, Vassilis Tsotras, Rares Vernica, Jian Wen, Till WestmannKnowledge-Based Trust: A Method to Estimate the Trustworthiness of Web SourcesXin Luna Dong, Evgeniy Gabrilovich, Kevin Murphy, Van Dang, Wilko Horn, Camillo Lugaresi, Shaohua Sun, Wei ZhangEfficient Evaluation of Object-Centric Exploration Queries for VisualizationYou Wu, Boulos Harb, Jun Yang, Cong YuInterpretable and Informative Explanations of OutcomesKareem El Gebaly, Parag Agrawal, Lukasz Golab, Flip Korn, Divesh SrivastavaTake me to your leader! Online Optimization of Distributed Storage ConfigurationsArtyom Sharov, Alexander Shraer, Arif Merchant, Murray StokelyTreeScope: Finding Structural Anomalies In Semi-Structured DataShanshan Ying, Flip Korn, Barna Saha, Divesh SrivastavaWorkshops:Workshop co-chair: Cong YuProgram committee includes: Sandeep TataInvited talk at BIRTE by: Ashish Gupta, Jeff ShuteDemonstrations:Xu Chu, John Morcos, Ihab Ilyas, Mourad Ouzzani, Paolo Papotti, Nan Tang, Yin YeXiaolan Wang, Mary Feng, Yue Wang, Xin Luna Dong, Alexandra Meliou

Announcing Google’s 2015 Global PhD Fellows

Friday, August 28, 2015

Posted by Michael Rennaker, Google University RelationsPhD Fellowship programAustraliaChina and East AsiaIndiaNorth AmericaEurope and the Middle EastAustralia

Bahar Salehi, Natural Language Processing (University of Melbourne)

Siqi Liu, Computational Neuroscience (University of Sydney)

Qian Ge, Systems (University of New South Wales)

China and East Asia

Bo Xin, Artificial Intelligence (Peking University)

Xingyu Zeng, Computer Vision (The Chinese University of Hong Kong)

Suining He, Mobile Computing (The Hong Kong University of Science and Technology)

Zhenzhe Zheng, Mobile Networking (Shanghai Jiao Tong University)

Jinpeng Wang, Natural Language Processing (Peking University)

Zijia Lin, Search and Information Retrieval (Tsinghua University)

Shinae Woo, Networking and Distributed Systems (Korea Advanced Institute of Science and Technology)

Jungdam Won, Robotics (Seoul National University)

India

Palash Dey, Algorithms (Indian Institute of Science)

Avisek Lahiri, Machine Perception (Indian Institute of Technology Kharagpur)

Malavika Samak, Programming Languages and Software Engineering (Indian Institute of Science)

Europe and the Middle East

Heike Adel, Natural Language Processing (University of Munich)

Thang Bui, Speech Technology (University of Cambridge)

Victoria Caparrós Cabezas, Distributed Systems (ETH Zurich)

Nadav Cohen, Machine Learning (The Hebrew University of Jerusalem)

Josip Djolonga, Probabilistic Inference (ETH Zurich)

Jakob Julian Engel, Computer Vision (Technische Universität München)

Nikola Gvozdiev, Computer Networking (University College London)

Felix Hill, Language Understanding (University of Cambridge)

Durk Kingma, Deep Learning (University of Amsterdam)

Massimo Nicosia, Statistical Natural Language Processing (University of Trento)

George Prekas, Operating Systems (École Polytechnique Fédérale de Lausanne)

Roman Prutkin, Graph Algorithms (Karlsruhe Institute of Technology)

Siva Reddy, Multilingual Semantic Parsing (The University of Edinburgh)

Immanuel Trummer, Structured Data Analysis (École Polytechnique Fédérale de Lausanne)

Margarita Vald, Security (Tel Aviv University)

North America

Waleed Ammar, Natural Language Processing (Carnegie Mellon University)

Justin Meza, Systems Reliability (Carnegie Mellon University)

Nick Arnosti, Market Algorithms (Stanford University)

Osbert Bastani, Programming Languages (Stanford University)

Saurabh Gupta, Computer Vision (University of California, Berkeley)

Masoud Moshref Javadi, Computer Networking (University of Southern California)

Muhammad Naveed, Security (University of Illinois at Urbana-Champaign)

Aaron Parks, Mobile Networking (University of Washington)

Kyle Rector, Human Computer Interaction (University of Washington)

Riley Spahn, Privacy (Columbia University)

Yun Teng, Computer Graphics (University of California, Santa Barbara)

Carl Vondrick, Machine Perception, (Massachusetts Institute of Technology)

Xiaolan Wang, Structured Data (University of Massachusetts Amherst)

Tan Zhang, Mobile Systems (University of Wisconsin-Madison)

Wojciech Zaremba, Machine Learning (New York University)

Google Faculty Research Awards: Summer 2015

Friday, August 21, 2015

posted by Maggie Johnson, Director of Education and University RelationsGoogle Faculty Research Awardslast roundrecipients of this round’s awardsour website

The Next Chapter for Flu Trends

Thursday, August 20, 2015

Posted by The Flu Trends Teamlaunchuseful insightssearch trendseconomicsother fieldsupdatingColumbia University’s Mailman School of Public HealthdashboardBoston Children’s Hospital/HarvardCenters for Disease Control and Prevention (CDC) Influenza Division.affect millions of people every yearFlu Trends web page

Pulling Back the Curtain on Google’s Network Infrastructure

Tuesday, August 18, 2015

Posted by Amin Vahdat, Google Fellowwarehouse-scale computerACM SIGCOMM conferencepapertalkOpen Network Summitbisection bandwidthrecent paperswitchesSoftware Defined Networking (SDN)

Our work on Bandwidth Enforcer shows how we can allocate wide area bandwidth among tens of thousands of individual applications based on centrally configured policy, substantially improving network utilization while simultaneously isolating services from one another.

Condor addresses the challenges of designing data center network topologies. Network designers can specify constraints for data center networks; Condor efficiently generates candidate network designs that meet these constraints, and evaluates these candidates against a variety of target metrics.

Congestion control in datacenter networks is challenging because of tiny buffers and very small round trip times. TIMELY shows how to manage datacenter bandwidth allocation while maintaining highly responsive and low latency network roundtrips in the data center.

Google Cloud Platform

Say hello to the Enigma conference

Tuesday, August 18, 2015

Posted by Elie Bursztein - Anti-abuse team, Parisa Tabriz - Chrome Security and Niels Provos - Security teamUSENIX Enigma

http://enigma.usenix.org

KDD 2015 Best Research Paper Award: “Algorithms for Public-Private Social Networks”

Monday, August 17, 2015

Posted by Posted by Corinna Cortes, Head, Google Research NY21st ACM conference on Knowledge Discovery and Data Miningbluefreely availableACM Digital LibraryEfficient Algorithms for Public-Private Social NetworksRavi KumarSilvio LattanziVahab MirrokniAlessandro EpastoFlavio ChierichettiBest Research PaperedgesIn a recent studypublic-privategraphssketchingsamplingsketchingPageRanksamplingKDD’15 Papers, co-authored by Googlers: Efficient Algorithms for Public-Private Social Networks(Best Paper Award)Flavio Chierichetti, Alessandro Epasto, Ravi Kumar, Silvio Lattanzi, Vahab MirrokniLarge-Scale Distributed Bayesian Matrix Factorization using Stochastic Gradient MCMCSungjin Ahn, Anoop Korattikara, Nathan Liu, Suju Rajan, Max WellingTimeMachine: Timeline Generation for Knowledge-Base EntitiesTim Althoff, Xin Luna Dong, Kevin Murphy, Safa Alai, Van Dang, Wei ZhangAlgorithmic Cartography: Placing Points of Interest and Ads on MapsMohammad Mahdian, Okke Schrijvers, Sergei VassilvitskiiStream Sampling for Frequency Cap StatisticsEdith CohenDirichlet-Hawkes Processes with Applications to Clustering Continuous-Time Document StreamsNan Du, Mehrdad Farajtabar, Amr Ahmed, Alexander J.Smola, Le SongAdaptation Algorithm and Theory Based on Generalized DiscrepancyCorinna Cortes, Mehryar Mohri, Andrés Muñoz Medina (now at Google)Estimating Local Intrinsic DimensionalityLaurent Amsaleg, Oussama Chelly, Teddy Furon, Stéphane Girard, Michael E. Houle Ken-ichi Kawarabayashi, Michael NettUnified and Contrasting Cuts in Multiple Graphs: Application to Medical Imaging SegmentationChia-Tung Kuo, Xiang Wang, Peter Walker, Owen Carmichael, Jieping Ye, Ian DavidsonGoing In-depth: Finding Longform on the WebVirginia Smith, Miriam Connor, Isabelle StantonAnnotating needles in the haystack without looking: Product information extraction from emailsWeinan Zhang, Amr Ahmed, Jie Yang, Vanja Josifovski, Alexander SmolaFocusing on the Long-term: It's Good for Users and BusinessDiane Tang, Henning Hohnhold, Deirdre O'Brien

Google’s Course Builder 1.9 improves instructor experience and takes Skill Maps to the next level

Thursday, August 13, 2015

Posted by Adam Feldman, Product Manager and Pavel Simakov, Technical Lead, Course Builder Team(Cross-posted on the Google for Education Blog)Course Builderskill mapping capabilitiesdownload it hereGoogle Open Online Education

Measuring competence with skill maps
In addition to defining skills and prerequisites for each lesson, you can now apply skills to each question in your courses’ assessments. By completing the assessments and activities, learners will be able to measure their level of competence for each skill. For instance, here’s what a student taking Power Searching with Google might see:

Improving usability when creating a course Course Builder has a rich set of capabilities, giving you control over every aspect of your course -- but that doesn’t mean it has to be hard to use. Our goal is to help you spend less time setting up your course and more time educating your students. We’ve completely reorganized the dashboard, reducing the number of tabs and making the settings you need clearer and easier to find.

release noteslet us knowSesame Street’s Make Believe with MathComputational Thinking for Educators

The neural networks behind Google Voice transcription

Tuesday, August 11, 2015

Posted by Françoise Beaufays, Research Scientistdeep learningimage classification and captioningtranslationmodel visualization techniqueswe announcedLong Short-term Memory Recurrent Neural NetworksGaussian Mixture Modeladapting the modelsrevolutionized the field of speech recognitiondifferentiating phonetic unitsfirst launched

An LSTM memory cell, showing the gating mechanisms that allow it to store
and communicate information. Image credit: Jürgen Schmidhuber

garbage in, garbage out

The reusable holdout: Preserving validity in adaptive data analysis

Thursday, August 6, 2015

Posted by Moritz Hardt, Research ScientistXKCD cartoonp-value

Image credit: XKCD

fixedadaptiveThe Reusable Holdout: Preserving Validity in Adaptive Data AnalysisCynthia DworkVitaly FeldmanToniann PitassiOmer ReingoldAaron RothSciencereusable holdoutThe curse of adaptivityFreedman’s paradoxF-test

Inference after selection: We first select a subset of the variables based on a data-dependent criterion and then fit a linear model on the selected variables.

“garden of the forking paths”Machine learning competitions and holdout sets

Reusable holdout setspre-registrationreusable holdout method

A general methodologydifferential privacystability

Example of a stable learning algorithm: Deletion of any single data point does not affect the accuracy of the classifier much.

generalizationadaptiveReliable benchmarksrelated work with Avrim Blumthis blog postLadder algorithmConclusion

Young people who are changing the world through science

Tuesday, August 4, 2015

Posted by Andrea Cohan, Google Science Fair Program Manager (Cross-posted from the Google for Education Blog)Google Science Fairannounce the 20 Global FinalistsSpotlight on a Young Scientist series on the Google for Education blog

a panel of notable international scientists and scholarsother incredible prizes

See through the clouds with Earth Engine and Sentinel-1 Data

Monday, August 3, 2015

Posted by Luc Vincent, Engineering Director, Geo ImageryGoogle Earth EngineEuropean Geosciences Union General AssemblyIEEE Geoscience and Remote Sensing Society

Noel Gorelick presenting Google Earth Engine at EGU 2015.

European Commission Joint Research CentreWageningen UniversityUniversity of Paviautilizing the Earth Engine geospatial analysis platformCopernicus Sentinel-1Copernicus

Sentinel-1 data visualized using Earth Engine, showing Vienna (left) and Milan (right).

Wind farms seen off the Eastern coast of England.

Blog