Probabilistic Cardinal Direction Queries On Spatio-Temporal Data Ganesh Viswanathan Midterm Project Report CIS 6930 Data Science: Large-Scale Advanced.

Slides:



Advertisements
Similar presentations
1 GOES-R Hurricane Intensity Estimation (HIE) Validation Tool Development Winds Application Team Tim Olander (CIMSS) Jaime Daniels (STAR)
Advertisements

An Interactive-Voting Based Map Matching Algorithm
GEOGRAPHIC INFORMATION SYSTEMS PRESENTATION 1
Ranking Outliers Using Symmetric Neighborhood Relationship Wen Jin, Anthony K.H. Tung, Jiawei Han, and Wei Wang Advances in Knowledge Discovery and Data.
Spatial Database Systems. Spatial Database Applications GIS applications (maps): Urban planning, route optimization, fire or pollution monitoring, utility.
Quantification of Spatially Distributed Errors of Precipitation Rates and Types from the TRMM Precipitation Radar 2A25 (the latest successive V6 and V7)
1 Chapter 5 : Query Processing and Optimization Group 4: Nipun Garg, Surabhi Mithal
Query Optimization of Frequent Itemset Mining on Multiple Databases Mining on Multiple Databases David Fuhry Department of Computer Science Kent State.
Tracking Tropical Cyclones in the Northwest Pacific During El Nino Andy Chan Environmental and Water Resources Engineering University of Texas at Austin.
Reported by Sujing Wang UH-DMML Group Meeting Nov. 22, 2010.
US Hurricanes and economic damage: an extreme value perspective Nick Cavanaugh, futurologist Dan Chavas, tempestologist Christina Karamperidou, statsinator.
Similarity Search in High Dimensions via Hashing
Patch to the Future: Unsupervised Visual Prediction
Development of a Web-based, Multimedia Database for Collection, Organization and Analysis of Biomedical Signals M.S.C.S. Problem Report Defense Lan Guo.
Spatial Mining.
Hechen Liu & Markus Schneider Department of Computer and Information Science and Engineering University of Florida Balloon: Representing and Querying the.
--Presented By Sudheer Chelluboina. Professor: Dr.Maggie Dunham.
CENTRE Cellular Network’s Positioning Data Generator Fosca GiannottiKDD-Lab Andrea MazzoniKKD-Lab Puntoni SimoneKDD-Lab Chiara RensoKDD-Lab.
Spatio-temporal Databases Time Parameterized Queries.
Spatio-Temporal Databases
Introduction This project deals with conversion of Vector based Probable Maximum Precipitation (PMP) data into Raster based PMP data using different interpolation.
Spatio-Temporal Databases. Outline Spatial Databases Temporal Databases Spatio-temporal Databases Multimedia Databases …..
Spatiotemporal GIS: Incorporating Time Group 7 Nathan Hunstad, Kyle Martin Csci 5980.
1 SINA: Scalable Incremental Processing of Continuous Queries in Spatio-temporal Databases Mohamed F. Mokbel, Xiaopeng Xiong, Walid G. Aref Presented by.
1 Location Information Management and Moving Object Databases “Moving Object Databases: Issues and Solutions” Ouri, Bo, Sam and Liqin.
1 SINA: Scalable Incremental Processing of Continuous Queries in Spatio-temporal Databases Mohamed F. Mokbel, Xiaopeng Xiong, Walid G. Aref Presented by.
CIS607, Fall 2005 Semantic Information Integration Article Name: Clio Grows Up: From Research Prototype to Industrial Tool Name: DH(Dong Hwi) kwak Date:
Analysis of High Resolution Infrared Images of Hurricanes from Polar Satellites as a Proxy for GOES-R INTRODUCTION GOES-R will include the Advanced Baseline.
National Weather Service – Newport/Morehead City NC NHC/WFO Tropical Products…and What’s New for 2012 WFO Newport Hurricane Awareness Seminar July 17,
Integrating Multi-Media with Geographical Information in the BORG Architecture R. George Department of Computer Science Clark Atlanta University Atlanta,
GeoPKDD Geographic Privacy-aware Knowledge Discovery and Delivery Kick-off meeting Pisa, March 14, 2005.
Title: Spatial Data Mining in Geo-Business. Overview  Twisting the Perspective of Map Surfaces — describes the character of spatial distributions through.
The Land Around Us An Introduction to Maps By: Mrs. Miles.
A Sweeper Scheme for Localization and Mobility Prediction in Underwater Acoustic Sensor Networks K. T. DharanC. Srimathi*Soo-Hyun Park VIT University Vellore,
Table of Contents Air Masses and Fronts Storms Predicting the Weather.
Improvements in Deterministic and Probabilistic Tropical Cyclone Wind Predictions: A Joint Hurricane Testbed Project Update Mark DeMaria and Ray Zehr NOAA/NESDIS/ORA,
1 A Bayesian Method for Guessing the Extreme Values in a Data Set Mingxi Wu, Chris Jermaine University of Florida September 2007.
Highline Class, BI 348 Basic Business Analytics using Excel, Chapter 01 Intro to Business Analytics BI 348, Chapter 01.
Update on Storm Surge at NCEP Dr. Rick Knabb, Director, National Hurricane Center and representing numerous partners 21 January 2014.
Hurricane lecture for KMA Ed Szoke 1 October 20, 2004 Overview of 2004 Atlantic Hurricane Season Ed Szoke* NOAA Forecast Systems Laboratory Forecast Research.
Spatial Data Analysis Yaji Sripada. Dept. of Computing Science, University of Aberdeen2 In this lecture you learn What is spatial data and their special.
Guidance on Intensity Guidance Kieran Bhatia, David Nolan, Mark DeMaria, Andrea Schumacher IHC Presentation This project is supported by the.
Pablo Santos WFO Miami, FL Mark DeMaria NOAA/NESDIS David Sharp WFO Melbourne, FL rd IHC St Petersburg, FL PS/DS “HURRICANE CONDITIONS EXPECTED.”
Astro / Geo / Eco - Sciences Illustrative examples of success stories: Sloan digital sky survey: data portal for astronomy data, 1M+ users and nearly 1B.
Extending Spatial Hot Spot Detection Techniques to Temporal Dimensions Sungsoon Hwang Department of Geography State University of New York at Buffalo DMGIS.
On the ability of global Ensemble Prediction Systems to predict tropical cyclone track probabilities Sharanya J. Majumdar and Peter M. Finocchio RSMAS.
Data Mining – Intro. Course Overview Spatial Databases Temporal and Spatio-Temporal Databases Multimedia Databases Data Mining.
Advanced Database Course (ESED5204) Eng. Hanan Alyazji University of Palestine Software Engineering Department.
Spatio-temporal Pattern Queries M. Hadjieleftheriou G. Kollios P. Bakalov V. J. Tsotras.
PREDICTABILITY OF WESTERN NORTH PACIFIC TROPICAL CYCLONE EVENTS ON INTRASEASONAL TIMESCALES WITH THE ECMWF MONTHLY FORECAST MODEL Russell L. Elsberry and.
Challenges in Mining Large Image Datasets Jelena Tešić, B.S. Manjunath University of California, Santa Barbara
Hurricane Investigation. Where Do Hurricanes Form? Purpose: Purpose: To determine the best location for scientific instruments designed to detect hurricanes.
A. FY12-13 GIMPAP Project Proposal Title Page version 26 October 2011 Title: Developing GOES-Based Tropical Cyclone Recurvature Tools Status: New Duration:
Exploitation of Ensemble Prediction System in support of Atlantic Tropical Cyclogenesis Prediction Chris Thorncroft Department of Atmospheric and Environmental.
Forecasting Storm Tracks Before Formation and Wind Speed per Platform Scott Morris Earth Science Associates Long Beach, CA November 5, 2015 User Conference.
1 86 th Annual American Meteorological Society Meeting Atlanta, Georgia January 29 – February 2, 2006 The Severe Weather Data Inventory (SWDI): A Geospatial.
64th Interdepartmental Hurricane Conference NOAA Tropical Program Delivering on the Promise of Partnerships Jack Hayes NOAA Assistant Administrator & Director,
GIS in Weather Instructor: Professor, Dr Yuji Murayama Teaching Assistant: Niloofar Haji Mirza Aghasi.
How are hurricanes formed ? A hurricane is a system of low pressure that originates over a tropical area. There are specific conditions that must be present.
Topics  Direct Predicate Characterization as an evaluation method.  Implementation and Testing of the Approach.  Conclusions and Future Work.
6TH Grade Science Notebook
A Visualization Tool for fMRI Data Mining
Sea State Influence on SFMR Wind Speed Retrievals in Tropical Cyclones
Clustering Uncertain Taxi data
Spatio-temporal Pattern Queries
Probabilistic Data Management
Unit 3 Typhoon Who? Please refer to page 18.
Table of Contents Air Masses and Fronts Storms Predicting the Weather.
Time Relaxed Spatiotemporal Trajectory Joins
Physics-guided machine learning for milling stability:
Presentation transcript:

Probabilistic Cardinal Direction Queries On Spatio-Temporal Data Ganesh Viswanathan Midterm Project Report CIS 6930 Data Science: Large-Scale Advanced Data Analytics University of Florida September 3 rd, 2011

Outline Introduction Uncertainty in spatio-temporal data Advanced queries on spatio-temporal data Cardinal direction relations (CDR) Probabilistic CDR Project Goals Methodology and analysis – what has been done – timeline for the future Conclusions

Uncertainty in Spatio-Temporal Data Systems for continuous monitoring or tracking of mobile objects receive updated locations of objects as they move in space Limitations of the bandwidth and battery power of mobile devices, make it infeasible for tracking the movement of objects with 100% certainty Example: If there is a time delay between capture of location and its insertion in the database, location values received by object may be different from actual locations In GIS, the root-mean-square-error (RMSE) is one approach to report this positional (in)accuracy

Advanced Queries on Spatio-Temporal Data Spatial relations can be Topological, Distance or Direction based Nearest-neighbor (NN), distance-range and direction-relation queries are important query types in spatial databases Probabilistic version of these advanced queries can speed up similarity joins among spatial relations disjoint contains inside equal meet covers coveredBy overlap 1 Km AB B lies to the East of A

Applications Applications in GIS, Cognitive Sciences, AI, Robotics, Qualitative spatial reasoning, density-based data mining techniques. In weather event analysis, probabilistic approaches can be used to improve the performance of join processing over large relations that contain moving object trajectories, to model the positional uncertainty of the moving eye of the hurricane

Project Goals 1.Query the trajectory of a hurricane to determine the direction taken by it at any instant t during its lifetime 2.Incorporate uncertainty: Enable probabilistic direction-relation queries among the spatio-temporal objects 3.Provide a visualization for the results based on tropical weather event data Example: Given objects O 1 and O 2 evaluate dir( ) and return a set of tuples of the form (O 1, O 2, d, p d ) such that p d is the probability of occurrence of the cardinal direction d between O 1 and O 2

Cardinal Direction Relations Besides its application in wayfinding, direction relationships are used in spatial databases and GIS as selection and join criteria in queries. Given two objects A and B, a function dir t (A,B) yields the direction relation of A w.r.t B at time t. Cardinal directions is an important qualitative spatial concept Direction relations Absolute (North, South, East, West, etc.) Relative (front, behind, left, right, etc.)

Cardinal Direction Relations Objects interaction grid (OIG) for direction finding A B

Cardinal Direction Relations Objects interaction grid (OIG) for direction finding OIG(A,B) = A B

Cardinal Direction Relations Interpretation A B 1. Determine the location of each component of object A & object B 2. Determine cardinal directions between the components

Probabilistic Cardinal Direction Relations Useful in performing similarity join queries Useful for positionally uncertain moving objects Probability of the direction between the tropical cyclone event at current location(s) and the location(s) at the next subsequent time instant Allows to leverage predictive models for forecasting the trajectory of newer storms and hurricanes based on previous patterns

Methodology and Analysis Steps involved Study of Related Work Data Collection Extensions to OIM for Probabilistic Direction Querying (PDQ) Predictive analysis of weather events using the probabilities, based on top- k or thresholding Visualization for PDQ results Experiments

Data Collection Best-track tropical weather information is obtained from three sources: – National Hurricane Center (NHC) – the National Oceanic and Atmospheric Administration (NOAA) – the Joint Typhoon Warning Center (JTWC) These datasets contain over 120k rows accounting for the spatio-temporal variation of tropical storm and hurricane events over the continental United States from 1990 to Spatial data for map boundaries of Continents, Counties, States, Counties and City locations obtained from data.gov All data has been downloaded, files parsed and converted into normalized database tables DONE!

Uncertainty Model and Probabilistic Queries

Enabling probabilistic direction relation queries on spatio-temporal data: Evaluation Idea t1t1 t2t2

Enabling probabilistic direction relation queries on spatio-temporal data: Evaluation Idea t1t1 t2t2

Enabling probabilistic direction relation queries on spatio-temporal data: Evaluation Idea Closed objects-interaction grid UAUA Tiling & OIM generation Interpretation UBUB for all t Generate probabilities for each & update database

1.Data collection – NHC, NOAA and JTWC hurricane data obtained and loaded into Oracle database (done) 2.Performing cardinal direction queries on spatio-temporal data (done) 3.Generation of direction pdfs for NHC, NOAA and JTWC datasets 4.Implementation of Probabilistic Direction Query (PDQ) algorithm 5.Testing and experiment analysis 6.Visualization using Google Maps API (partly done) Timeline Data Collection (Tropical weather event information) Generation of direction pdfs for NHC, NOAA and JTWC datasets Implementation of pOIM Visualization and Experiments

CONCLUSIONS The work studies probabilistic queries on spatio-temporal data and defines a novel query type: probabilistic cardinal direction query on them Illustrates a large-scale data science application for using probabilistic cardinal direction querying to improve weather event analysis Future work can include: Extensions of probabilistic Nearest Neighbor queries using both distance and direction, testing of similarity joins with PDQ and exploration of probabilistic topological querying operations on uncertain data. Questions?