News

• Scaling deep learning training and inference via model-wise distributed memory parallelism and hypergraph partitioning, ACM ICS '21
• Private spatio-temporal trajectory sharing with local differential privacy, VLDB '21
• Spatio-temporal quantization for compressed storage and querying of large datasets, VLDB '21
• Synthetic spatial data generation with differential privacy, SSTD '21
• Project on spatio-temporal graph analytics, with WMCA and ADEPT Live Labs; post-doctoral researcher position
• Best CS dissertation award to Sean Hooker for his work in our group on sparsification and quantization of deep neural networks
• Project on spatio-temporal ML, with TfWM, Immense Simulations, one.network, blacc, and O2 (Telefonica UK)
• Spatio-temporal forecasting via graph ML, ACM IMWUT '20
• Scalable linkage for large-scale spatio-temporal databases, ACM SIGMOD '20

Biography

Hakan Ferhatosmanoglu is a Professor at the University of Warwick, and a Fellow at the Alan Turing Institute, the UK's national institute for data science and AI. His research interest is in building scalable data systems, large-scale search and data analytics. The research problems include efficient management of information from multiple sources while protecting the privacy of data subjects. An example application is to learn aggregate mobility patterns over large spatio-temporal databases. He has published in the premier venues of data systems and analytics, including VLDB, KDD, SIGMOD, PODS, ICDE, SIGIR, CIKM, WSDM, ICS, VLDBJ, TKDE, and scalable data science solutions at IEEE T. Mobile Computing, Mobile Data Management, UbiComp. He has been the principal investigator of more than 20 externally funded projects, and closely working with industrial partners in developing scalable machine learning solutions and data analytics systems.

Ferhatosmanoglu received his PhD in CS from University of California, Santa Barbara in 2001. He worked as a tenured professor at Ohio State U. (OSU), Bilkent U., and visited Max Planck Institute (MPI) for Informatics and MPI for Software Systems with a research award by the Humboldt Foundation. Past advisees work in permanent faculty positions at major research universities, and technology companies including Amazon, Facebook, Google, Microsoft, Oracle, Teradata. He received several prestigious Research Career Awards, including from the US Department of Energy, US National Science Foundation, and Alexander von Humboldt Foundation. He is currently the Director of Computer Science Centre for Doctoral Training and Research, Director of EPSRC CDT on Urban Science and Progress, Deputy Head of Department of Computer Science, and member of the Board of Faculty of Science, Engineering, and Medicine at the University of Warwick.

Program/Organization Committee

SIGMOD 2022, 2021, 2018; ICDE 2022, 2021, 2020, 2019; Mobile Data Management 2021, 2020, 2019; CIKM 2021, 2020, 2019, 2017; VLDB 2019, 2018, 2017; SIGIR 2017, 2016; SIGMOD 2016, VLDB 2016, ECML/PKDD 2015, ECML/PKDD 2014, VLDB 2014, VLDB 2013, VLDB 2012 (Co-Chair, Workshops Program), VLDB 2012 (PC), SIGMOD 2011, ICDE 2011, ICDM 2011, ICDM 2010, ACM SIGSPATIAL 2010, SIGMOD 2010 (Org.), SIGMOD 2010 (PC)...

Recent Publications

Scalable Data Systems and Analytics

• Partitioning Sparse Deep Neural Networks for Scalable Training and Inference
• SLIM: Scalable Linkage of Mobility Data
• Cascade-aware Partitioning of Large Graph Databases
• Distributed Block Formation and Layout for Disk-based Management of Large-Scale Graphs
• Topic-Based Influence Computation in Social Networks under Resource Constraints
• VISIR: Visual and Semantic Image Label Refinement
• Scaling Forecasting Algorithms using Clustered Modeling
• ALACA: A Platform for Dynamic Alarm Collection and Alert Notification in Network Management Systems
• Learning to Un-rank: Quantifying Search Exposure for Users in Online Communities
• Aggregate Profile Clustering for Telco Analytics
• Temporal Workload-Aware Replicated Partitioning for Social Networks
• λ-diverse Nearest Neighbors Browsing for Multi-dimensional data
• S3-TM: Scalable Streaming Short Text Matching

Spatial-Temporal Data Management

• Real-World Trajectory Sharing with Local Differential Privacy
• Privacy-Preserving Synthetic Location Data in the Real World
• PPQ-Trajectory: Spatio-temporal Quantization for Querying in Large Trajectory Repositories
• Spatio-Temporal Linkage over Location-Enhanced Services
• Fair Task Allocation in Crowdsourced Delivery
• Privacy-Preserving Aggregate Queries for Optimal Location Selection
• D3P: Data-driven Demand Prediction for Fast Expanding Electric Vehicle Sharing Systems
• Diverse Relevance Feedback for Time Series with Autoencoder Based Summarizations
• Predicting Optimal Facility Location Without Customer Locations
• Generating Time-Varying Road Network Data Using Sparse Trajectories
• Multi-Vehicle Route Planning for Efficient Urban Freight Transport
• Div-BLAST: Diversification of Sequence Search Results
• LineKing: Coffee Shop Wait-Time Monitoring Using Smartphones
• Diversity based Relevance Feedback for Time Series Search