Jianting Zhang's Picture

Jianting Zhang

Assistant professor in Geographical Information System (GIS)


Affilications:
CS@CUNY City College (Primary)

CS@CUNY Graduate CenterNOAA-CRESTUTRC,
CCSI@ORNL (through DOE VFP)


Geospatial Technologies and Environmental Cyberinfrastructure (GeoTECI) Lab


NSF IIS-Medium Collaborative: Spaital Data and Trajectory Data Managment on GPUs



Facutly Profile[Education/Training] [Professional Experiences] [Courses] [Contact Info][Archieved News]

GeoTECI Lab [Overview (9 slides)] [Students] [Hardware] [Software][Data] [Research Code and Online Demos]

Recent News:
[08/20/2014] A brief introduction to my research and my group (GeoTECI lab) with 9 slides is added.

[11/05/2014] GeoTECI Ph.D. student Simin You has pre-released two Big Spatial Data Managment prototype systems: SpatialSpark based on Apache Spark and ISP based Cloudera Impala. ISP includes both a multi-core CPU implementaiton and  a GPU implementation by signficantly extending Impala. Both SpatialSpark and ISP are based our data parallel designs for spatial data processing that were inititally developed for GPUs for single comuting nodes (please see the publication section for details). While we are still in the process of assembling codes and writting technical reports, preliminary results on two real world datasets (~170 million NYC Taxi trips and ~375 million GBIF species occurences) have shown 10-100X performance speedups over those based on Hadoop/MapReduce. Our experiments also releaved the arctitectural inefficiencies of Hadoop/MaprReduce in processing semi-structured data (non-relational data in general and spatila data in particular)  in distributed environments. We are in the process of developing prototypes directly on top of HDFS and possibly additional open source distributed file systems. Please contact me and Simin for early accesses to source code and/or Amazon EC2 images if interested. 


Research Interests

·        High-Performance Geospatial Computing (HPC-G)

·        Spatial Databases (SDB) and GIS Applications (GIS)

·        Environmental Cyberinfrastructure (CI)

·        Geospatial Visual Analytics (GVA)

·        Multispectral and Hyperspectral Remote Sensing Data Processing (RS)

Publication (DBLP)

        ·        By Topics (Selected): HPC-G, SDB, GIS, RS, GVA, CI

        ·        By Year: 2014, 2013, 2012, 2011, 2010 2009, 2008, 2007, 2006, 2005, 2004, 2003, 2002, 2001, 2000, 1999, 1998, 1997

 

Research Initiatives and Exploratory Projects [Funded Projects]

·   High-Performance Geospatial Computing on GPGPUs [Overview] (* Experimental Source Code)
(1) Previous Results: [HPC-G (Position Paper)] [BMMQ-Tree] [BQ-Tree] [NNI-DEM][Polygon Rasterization]

                        (2) Reference Primitives based Implementations [PrimQuad*][PrimCSPTP*][PrimTrQuery*] [PrimSpJoin*]
                        (3) Spatial Indexing and Query Optimization: [R-Tree][Selectivity Estimation]

                        (4) End-to-End Systems for Spatial Joins (10-40X over in-memory systems and 3-4 orders of speedups over disk-resident systems)
                             [Point to Network][Point in Polygon][Point to Polygon][Trajectory to Trajectory]

                        (5) In-Progress (Algorithms): [Indexing Polygon Internals] [Increamental Refining Filtering]  [Polygon to Polygon Join (Overlay)]
                        (6) In-Progress (Systems) [LLVM-based Spatial SQL Front-end] [Hybrid CPU-GPU Systems] [Scaling-Out to Clusters]

·     Managing Large-Scale  OD/Trajectory Data with Applications to Data Mining of Traffic and Travel Patterns [Overview]
[U2SOD-DB] [U2SOD-VA][NYC Case Studies] [Frequenet Sequence Mining using Shortcuts]

·     Managing Large-Scale Environmental and Species Distribution Data with Applications to Understanding Global and Regional Biodiversity Patterns  [Overview] [Visual Exploreations of Rasters] [Lightweight Compression for Rasters] [USGS Little Tree Range Map] [NatureServe Bird Range Map in DBMSs] [Zonal Summation of GBIF Data on GPUs]

·    Web-based Query-Driven Visual Exploration of Large-Scale Geospatial Data (*Online Accessible)
 
[Overview] [NYC CrashMap*][BirdsQuests*][RasterExplorer*] [HP-GVE]