Transcript Literature Search - COW :: Ceng

CENG 553 TERM PAPER PRESENTATION
Spatio-Temporal Stream Processing &
Data Aggregation Techniques
Cem UTAN
[email protected]
OUTLINE
•
•
•
•
•
•
•
•
Spatial Databases
SDBMS vs. GIS
Issues & Problems in SDBMS
Stream Applications vs. Spatio Temporal DBMS
Spatio-Temporal Analytics
Spatio Temporal Analytic Tools
Conclusions
References
Spatial Databases
Spatial Databases (SDBMS)
• Traditional (non-spatial) database management systems provide:
• Persistence across failures
• Allows concurrent access to data
• Scalability to search queries on very large datasets which do not fit inside
main memories of computers
• Efficient for non-spatial queries, but not for spatial queries
• Non-spatial queries:
• List the names of all bookstore with more than ten thousand titles.
• List the names of ten customers, in terms of sales, in the year 2001
• Use an index to narrow down the search
Spatial Databases (SDBMS)
• Spatial Queries:
• List the names of all bookstores with ten kilometers of Rio de Jenario
• List all customers who live in Sau Paulo and its adjoining districts
• List all the customers who reside within fifty kilometers of the company
headquarter
sample buffer analysis
Value of SDBMS
• Examples of non-spatial data
• Names, phone numbers, email addresses of people
• Examples of Spatial data
• Census Data
• NASA satellites imagery - terabytes of data per day
• Weather and Climate Data
• Rivers, Farms, ecological impact
• Medical Imaging
User Application Domains
• Many important application domains have spatial data and queries. Some
Examples follow:
• Army Field Commander: Has there been any significant enemy troop
movement since last night?
• Insurance Risk Manager: Which homes are most likely to be affected
in the next great flood on the Mississippi?
• Medical Doctor: Based on this patient's MRI, have we treated
somebody with a similar condition ?
• Molecular Biologist: Is the topology of the amino acid biosynthesis
gene in the genome found in any other sequence feature map in the
database ?
• Astronomer: Find all blue galaxies within 2 arming of quasars.
SDBMS
• A SDBMS is a software module that
• can work with an underlying DBMS
• supports spatial data models, spatial abstract data types (ADTs) and a
query language from which these ADTs are callable
• supports spatial indexing, efficient algorithms for processing spatial
operations, and domain specific rules for query optimization
• Example: Oracle Spatial data cartridge, ESRI SDE
• can work with Oracle 11g DBMS
• Has spatial data types (e.g. polygon), operations (e.g. overlap) callable
from SQL3 query language
• Has spatial indices, e.g. R-trees
• IBM: Spatial Option
• Informix: Spatial Data blade
SDBMS vs. GIS
SDBMS vs. GIS
• GIS is a software to visualize and analyze spatial data using spatial analysis
functions such as
Search Thematic search, search by region, (re-)classification
Location analysis Buffer, corridor, overlay
Terrain analysis Slope/aspect, catchment, drainage network
Flow analysis Connectivity, shortest path
Distribution Change detection, proximity, nearest neighbor
Spatial analysis/Statistics Pattern, centrality, autocorrelation, indices of
similarity, topology: hole description
• Measurements Distance, perimeter, shape, adjacency, direction
•
•
•
•
•
•
• GIS uses SDBMS
• to store, search, query, share large spatial data sets
SDBMS vs. GIS
• SDBMS focuses on
• Efficient storage, querying, sharing of large spatial datasets
• Provides simpler set based query operations
• Example operations: search by region, overlay, nearest neighbor,
distance, adjacency, perimeter etc.
• Uses spatial indices and query optimization to speedup queries over
large spatial datasets.
• SDBMS may be used by applications other than GIS
• Astronomy, Genomics, Multimedia information systems, ...
Volunteer Geographic Information
Systems
•
•
•
•
OpenStreetMap,
Wikimapia
Foursquare
Trapster
Spatio-temporal Databases
• Add temporal dimension to data sets captured from resources.
• Examples:
• Trajectories
• Evolving region
• Moving points
Sample spatio-temporal data[63]
Issues and Related Problems
in S-DBMS
Issues in SDBMS
•
•
•
•
•
•
Spatial data model
Query language
Query processing
File organization and indices
Query optimization,
Data aggregation, etc.
Spatial Data Model
The Polygonal Structure
• SDBMS stores spatial data
as if they are non-spatial
data.
Spatial Data Model and its Problems
•
SDBMS stores spatial data as if they are non-spatial data. Even OGC (Open
Geospatial Consortium) has defined basic structure of geospatial data, some
changes exist and change on Spatial data standard effects other standards
[52]
OGC
GIS
•
Heterogeneous multi-source data have not managed under a unified spatiotemporal model, thus the data and products cannot fuse properly for display.
Middleware application services work for that purpose. i.e. Web processing
services. etc. [77]
Query Language and its Problems
• Standard SQL queries are not sufficient enough to analyze spatial data sets
• Different DBMS systems have developed and added GIS library queries
into DBMS. SQL3 is developed for this purpose.
• Application domain masters developed their own efficient languages[53]
• Examples:
• Oracle SDO. (spatial data operators) major one. [85]
• Moving Object Monitoring Query Language (MQL) [53]
Query processing and its Problems
• Due to different type of data structure, query processing must be in a different
manner. Data structure seems to have sequence tables.
• Processing a query with a sequence operation relatively requires much
computation and memory. [54]
Raster
image
SQL
DB
Vector data
Sensor data
Query processing and its Problems
Most common type is K-nearest neighborhood query processing. [69] [73]
• Problem:Continuous K-nearest neighbor (CKNN) query: it is an
important type of spatio-temporal query. Due to the large amount of CKNN
queries that need to be evaluated concurrently, efficiently processing such
queries inevitably becomes more complicated. [69]
• Solution:Candidates-Indicating-tree: The goal of designing the
candidates-indicating-tree (CI-tree) is to rapidly determine the candidates
for each KNN query no matter where and when the query is issued. The idea
of constructing the CI-tree is to recursively partition the data space
[0, w]2 into four equi-width grid cells until the number of partitions is equal
to a user-defined value l. Then, their candidate sets within the time interval
[tc, tc + Δt] would be predetermined and stored in the CI-tree for query
processing. [69]
File Organizations and Indices
The indexing mechanism
is the R-tree structure. R means
‘’rectangles’’. The rectangles who
have more common values inside
should be in same level. This
means that at higher levels the
aggregation of an increasing
number of objects. This can also
be seen as an increasingly coarse
approximation of the data set. [84]
File Organizations&Indices and
Problems
•
Problem: Querying about the time-varying location of specific moving objects
can be particularly expensive in terms of running time. Alternative indexing mechanisms
are generated for several data retrieval types. Especially for a temporal database time
dimension is needed to be indexed. [66]
•
Solution ie. meta-index: The meta-index is an indexing data structure that guides the
search process, is stored separately from the servers data, and handles partial data
contained in database servers. A meta-index called statistics- and trace-based meta-index,
which is designed to solve time-slice or time interval queries, and queries about the
location of a particular object at a given time instant. [66]
Data Aggregation & Problems
• Problem: Data Types. As data type changes aggregate functions should change
due to metadata specifications. While non-spatial databases use sum, count, avg etc.
Aggregates on data spatial data needs to be manipulated over its physical properties.
i.e. Length of a line, area of a polygon, intersection of two rectangles etc. These type
of aggregates are first generated by GIS software and then DBMS started to include
these properties by using different functions.
• Problem: Aggregate Functions. Recently, many spatio-temporal applications
paid attention to the aggregative information of moving objects in networks (e.g., the
number of the vehicles, the average speed). Conventional methods can solve the
distinct counting problem, but can’t guarantee on the approximate quality of
aggregate queries over moving objects in road networks in masses of situations. [67]
Sample Aggregate Functions From
Oracle SDBMS [85]
SDO_GEOM.SDO_AREA
Computes the area of a two-dimensional polygon.
SDO_GEOM.SDO_BUFFER
Generates a buffer polygon around or inside a geometry.
SDO_GEOM.SDO_CENTROID
SDO_GEOM.SDO_CONVEXHULL
Returns the centroid of a polygon.
Returns a polygon-type object that represents the
convex hull of a geometry object.
Returns a geometry object that is the topological
difference (MINUS operation) of two geometry objects.
Computes the distance between two geometry objects.
SDO_GEOM.SDO_DIFFERENCE
SDO_GEOM.SDO_DISTANCE
SDO_GEOM.SDO_INTERSECTION
SDO_GEOM.SDO_LENGTH
Returns a geometry object that is the topological
intersection (AND operation) of two geometry objects.
Computes the length or perimeter of a geometry.
-- Return the areas of all cola markets.
SELECT name, SDO_GEOM.SDO_AREA(shape, 0.005) FROM cola_markets;
Stream Applications vs.
Spatio Temporal DBMS
Geo-streams
• Current streams systems lack native spatial support
• Spatial stream queries are common in
• traffic monitoring
• environment monitoring
• moving object databases
Geo-stream databases
• Many data are generated continuously
• Transaction data
• Network monitoring [67] Financial application
• Most recent data are commonly queried in a one-pass fashion
• Monitoring
• Aggregation
• Database system provides abstractions and declarative languages
that stream processing can benefit from
Stream Applications
• Environmental monitoring: Spatial databases have a major role to investigate
the nature of the world. They provide us a good understanding of general or
extraordinary behavior of the nature. So it was frequently used for geographic
sciences. [79] [76]
• Patient monitoring: Temporal data about some illnesses on people are observed
via thematic maps and the reasons are defined more clearly by using stream data. [80]
• Network monitoring: All network objects in earth can be tracked via these
applications. i.e. Google map, navigation data: [57] [53]
• Transaction monitoring:[57] [53]
• Environment monitoring [77]
Network monitoring
Stream Applications
• Traffic analysis:[57]
• Moving object queries: These types of temporal data applications
provide time varying data abstractions [66] [67] [73] [53]
• Sensor network: Sensor based data sources like GPS continuously
provides stream data into Moving Objects Databases. Many spatial
datasets, in both vector and raster format as well as derived from statistical
yearbooks and sensor and GPS measurements were integrated in the
database. [72]
Sample
Sensor
Network
Sample Applications
• Environmental monitoring
• Notify me when UV is high, temperature is low
• Traffic monitoring
• Traffic jam: aggregated speed much below speed limit on a road
segment for extended time
• Accident: vehicle on unintended space, e.g. high way for longer
than expected time
• Sensor Network
Take all images from satellite that have changed more recently.
Sample Applications
Vehicle tracking or Navigation applications
Route prediction
•
•
•
•
Next position
Next stop
The entire route
Application:
• Mobile commerce
• Save energy
• Traffic notification
Location-based social networking
• Social networking with location
•
•
•
•
•
Loopts
Google latitude
Geocache
Social dynamics
IPhone applications
A Spatial Stream Processing System
Architecture
Spatio Temporal Analytics
Spatio-temporal Analytics
• The analysis of data with both spatial and temporal
information
• The data are spatially and/or temporally correlated
"Everything is related to everything else, but near things are more related
than distant things. [86] Core idea behind R+tree
Why do we need spatio-temporal analytics?
• Analytics help us to describe what happened in the past,
understand what is happening now, predict what will
happen in the future, and make decisions.
• The proliferation of sensor devices makes spatio-temporal
information a fundamental component for almost every
analytical applications
Types of Spatio-Temporal Analytics Methods
•
•
•
•
•
•
Visualization and exploratory analysis
Outlier analysis
Segmentation (classification and clustering)
Colocation mining
Dependency analysis
Trend discovery
Data Visualization and Exploratory Analysis
• Map querying task :
• Query map data via its behavior can only be realized by visual tools. Google Earth
combines satellite raster imagery, with vector maps and layers, in a single and integrated
tool, which allows users to interactively fly in 3D from outer space to street level views. [55]
• New solutions do not only include a static graphical view of the results produced during
the data mining process, but also the possibility to dynamically and interactively obtain
different spatial and temporal views as well as interact in different ways with them [55]
Data Visualization and Exploratory Analysis
• Static query: (one-time query using map tools available on the
interface). a new map is not generated. Only objects are queried on
maps.
• Dynamic query[36]: (setup of event alert conditions). Generally a
change on the map is observed. New maps are generated. i.e. Thematic
maps.
• Spatial constraints are expressed using the map, while temporal
constraints are expressed as linear time moments[37]
Static query
[89]
Dynamic query[1]
Spatio-Temporal Outlier Analysis
• Definition of outliers: “Outliers can be defined as observations
which appear to be inconsistent with the remainder of the dataset.
They deviate too much from other observations. Outlier detection
is a data mining technique like classification, clustering, and
association rules.’’ [58]
outlier
Spatio-Temporal Outlier Analysis
• Spatial Outlier:
(S-Outlier) is an object
whose non-spatial
attribute value is
significantly different from
the values of its spatial
neighbors. [58]
Spatio-Temporal Outlier Analysis
Temporal Outlier :
(T-Outlier) is an object whose
non-spatial attribute value is
significantly different from
those of other objects in its
temporal neighborhood. The
studies [87] [88]
focus on the representation of a
time location which deviates
too much from its temporal
neighbors. [58]
Spatio-Temporal Outlier Analysis
• Spatio-Temporal Outlier:(ST-Outlier) is an object whose non-spatial
attribute value is significantly different from those of other objects in its
spatial and temporal neighborhood.
Spatio-Temporal Outlier Analysis
• Methods[48][58]
•
•
•
•
Clustering-based approach
Distance based approach
Computational geometry based approach
Spatial scan based approach
Segmentation methods
• Classification[41]
Spatial data is grouped via its certain attributes or geographic relations. [75]
• Spatial classification: decision tree, Bayesian, ANN…
• Temporal classification: decision tree, Bayesian, ANN…
• Temporal extensions to spatial classification/ Spatial extension to
temporal classification[75]
Segmentation methods
• Clustering:
A clustering-based map overlay approach treats every spatial attribute
as a map layer and considers spatial clusters (regions) of point-data in
each layer as candidates for mining associations[42] [57]
Segmentation methods:Clustering Types
• Spatial clustering: Partitioning method, hierarchical method, density based method,
and grid-based method.
• Temporal clustering: Cluster due to time slices.
• Interactive spatio-temporal clustering: Perform clustering spatially or temporally
and then test whether the cluster exist in both dimensions (EMM Test[43])
• Simultaneous spatio-temporal clustering: Space-time scan[44]
Spatio-temporal
clustering[44]
Segmentation methods: Clustering
• In traditional clustering the similarity measure often depends on the distance
between objects or their land surface types. The most popular clustering technique
is k-means. All these algorithms deal only with spatial correlations and cannot be
directly used for discovering temporal patterns simultaneously. These techniques
are very computationally expensive and usually trapped into local minima, as they
tend to concentrate on local features [55] [56]
Spatial
clustering [55]
Segmentation methods:Clustering techniques
• Density Based Clustering: Define a multivariate density distribution and
look for a set of fitting parameters for the model. It was proposed by
M.Breunig, et al. This method assigns a Local Outlier Factor (LOF) to each
sample based on their local neighborhood density. Samples with high LOF
value are identified as outliers. The neighborhood is defined by using Min Pts.
parameter [52][58]
• Distance-based method: Use a distance metric to measure the distances
among the data points. Problems may occur if the parameters of the data are
very different from each other indifferent regions of the data set. [58]
Density based clustering[84]
Distance Based clustering[60]
Segmentation methods:Clustering techniques
• Depth-based approaches are based on computational geometry and
compute different layers of k-d convex hulls. Outliers are more likely to be
data objects with smaller depths. However, in practice, this technique
becomes inefficient for large datasets (k>=4).Depth-based approach is also
applied for spatial outlier detection.[58][84]
• Distribution-based approaches use standard statistical distribution. They
deploy some standard distribution model (e.g. Normal, Poisson, etc.) and
recognize as outliers those points which deviate from the model. [58][84]
Distribution based clustering[84]
Segmentation methods: Other Clustering
techniques
• Model-based clustering [46]
• Moving object similarity search[72]
• DBSCAN extensions, OPTICS[47]
• Moving clusters[47][57]
•
Applications: movement data, cellular networks, environment
data…
Co-Location Mining
• Colocation mining finds subset of Boolean features located in
spatial proximity[60]
Co-Location Mining: Methods
•Data mining-based approach: Can be further divided into a clusteringbased map overlay approach and association rule-based approaches. [60]
•
A clustering-based map overlay approach treats every spatial
attribute as a map layer and considers spatial clusters (regions) of pointdata in each layer as candidates for mining associations.
• Association rule-based approaches can be divided into transactionbased approaches and distance-based approaches.
•Transaction based: Under this model, transactions are
created around instances of one user- special spatial feature.
The association rules are derived using the Apriori algorithm
•Distance-based : Patterns called k-neighboring class sets.
Co-Location Mining: Other Methods
•Spatial statistical approach[71]
•Buffer-based model
•Temporal extension: mixed-drove approach, weighted windowbased model[51]
Spatial statistics[71]
Dependency Analysis
• Dependency analysis: Commonly used in software maintenance to
understand the architectural interdependencies of software systems. For
example, an object-oriented system is composed of a collection of
communicating objects. Analyzing objects and their relationships
facilitates an in-depth understanding of object-oriented software systems.
For systems consisting of large quantities of objects, reverse engineering
tools can create package diagrams by grouping classes into packages. A
different approach aggregates the complete static description of software
objects so that each coarse-grained entity represents a set of objects. A
dependency between coarse grained entities indicates a possible
relationship between the objects they represent [52]
Other methods
• Association rule mining: Association rule mining seeks to discover
associations among transactions encoded in a database. An association rule
takes the form A→B where A(the antecedent) and B(the consequent) are sets
of predicates. A spatio-temporal association rule occurs when there is a
spatio-temporal relationship in the antecedent or consequent of the rule [59]
• Spatial preprocessing is required to discretize spatial measurements
• Methods[49]
• Bayesian networks
• Hieratical approach
• Trend discovery[81]
• Regression
• Sequence mining
Spatio Temporal Analytic Tools
List of Current Spatio-Temporal Analytics Tools
• Commercial
•
•
•
•
ESRI ArcGIS series
Microsoft SQL Spatial +Stream Insight
MapInfo[19]
Other commercial tools
• Open source/free software
• Descartes and Common GIS
• Map Server
• Other free tools
ESRI ArcGIS Series
• ArcGIS desktop and server provide most advanced and
complete toolkit
• Has many extensions for different domains
• Can use APIs to develop extensions, web or desktop
applications for customized needs. Many other commercial
tools such as CUBE[9] are built on top of ArcGIS.
ESRI ArcGIS Desktop and Server Extensions[1]
• 3D Extension (Desktop and Server)
• Analyze terrain data, model subsurface features, view and analyze
impact zones, determine optimum facility placement, share 3D
views, create a 3D virtual city.
• Geostatistical Extension (Desktop and Server)
• Visualize, model, and predict spatial relationships.
• Link data, graphs, and maps dynamically.
• Perform deterministic and geostatistical interpolation.
• Evaluate models and predictions probabilistically
ESRI ArcGIS Desktop and Server Extensions
• Network Extension (Desktop and Server)
• Dynamically model realistic network conditions and solve vehicle
routing problems
• Multipoint optimized routing, time-sensitive, turn-by-turn driving
directions , allocation of service areas, determining the fastest fixed
route to the closest facility
• Schematics Extension (Desktop and Server)
• Rapid checking of network connectivity
• Automatically generate schematics
Microsoft SQL Library + Stream Insight[2]
• Combines SQL Server spatial library with stream processing engine
• Integrating SQL library within Stream Insight engine
• Focuses on data stream event processing workflow
• GIS Support relies on SQL Server (limited), and therefore need
extensive customization for applications
Other Commercial tools by category
• Complete GIS Suite (similar to ArcGIS)
• Cardcorp SIS[8],Geomedia[17], IDRISI[18] ,
• Spatio-temporal analysis
• STIS[23]
• Network (traffic) analysis tools
• ACCESSION GIS[3], AltaMap Suite[4], CUBE[9], DYNAMEQ[15],
EMME[14]
• Terrain analysis
• ANUDEM[5]
• CAD applications
• AutoCAD Civil3D[6],
• Emergency and hazard modeling and analysis
• CadnaA[10], Calpuff View[11],Caris[12],CATS[13],Floodworks[16]
• Specialized analysis
• ClusterSeer and BoundarySeer[7] (cluster and boundary analysis),
Mathematica[20]
• Mathematics toolkit
• Matlab Plus toolbox[21], SPSS[22]
Descartes and CommonGIS[24]
• An interactive java based GIS tool for visualization and exploratory
analysis.
• Functionalities
• Map and graph visualization (Choropleth maps, scatter plot…)
• Basic queries (distance, difference…)
• Dynamic queries
• Open source and customizable, lack advanced GIS analytics
functionalities
MapServer[31]
• Open source GIS data rendering engine
• Functionalities
•
•
•
•
Advanced cartographic output
Cross platform and APIs for all popular scripting languages
Support many formats
OGC standard compliant
• Not a full GIS suite
Other Free/Open Source tools by category
• Complete GIS Suite (similar to ArcGIS)
• GRASS[28]
• Spatio-temporal analysis
• Map comparison kit[30], STAR[34]
• Terrain analysis
• Landserf[29]
• Exploratory data analysis
• GeoDA[26]
• Database extension
• PostGIS[32]
• Specialized analysis
• GAM/K[25](Clustering), GRASP[27](Regression)
• Mathematics toolkit
• R Spatial[33]
Conclusions
Detected Problems For Spatial & SpatioTemporal Databases


Sensor networks that are providing spatial data to spatial databases
are not common.
The heterogeneous and discrete information retrieval is a big
problem.

Spatial and non-spatial data sets are not linked properly.

Spatial data generation is an expensive process.

Spatial data management is a difficult job.

Performance Tuning is a big challenge on these data bases.

External GIS software is needed for data mining and visualization.


Topological relations are not provided quickly compared to flat
database relations.
3D data model in SDBMS is a lack of property.
Summing Up
• Spatio-temporal analytics is becoming a fundamental
component of business analytics.
• SDBMS has grown slower compared to GIS.
• SDBMS should include all spatial analytics located in GIS
software.
• Spatial data should be generated in a standard format.
• The future
• Big data (bigger due to spatio-temporal dimension)
• Real time (not only historical spatio-temporal data, but also streaming
data that requires optimization at all levels)
Thank You!
References (I)
[1] http://www.esri.com/software/arcgis/index.html
[2] http://msdn.microsoft.com/en-us/library/ee362541.aspx
[3] http://www.citilabs.com/accession.html
[4] http://www.geomicro.com/
[5] http://fennerschool.anu.edu.au/publications/software/anudem.php
[6] http://usa.autodesk.com/adsk/servlet/pc/index?siteID=123112&id=8777380
[7] http://www.terraseer.com/products_boundaryseer.php
[8] http://www.cadcorp.com/products_geographical_information_systems/index.htm
[9] http://www.citilabs.com/cube_base.html
[10] http://www.datakustik.com/en/products/cadnaa/
[11] http://www.weblakes.com/calpuff/calpuff_overview.html
[12] http://www.caris.com
[13] http://www.saic.com/products/security/cats/
[14] http://www.inro.ca/en/products/emme/
[15] http://www.inro.ca/en/products/dynameq/
[16] http://www.wallingfordsoftware.com/uk/products/floodworks/
[17] http://www.intergraph.com/sgi/default.aspx
[18] http://www.clarklabs.org/
[19] http://www.pbinsight.com/welcome/mapinfo/
[20] http://www.wolfram.com/products/mathematica/newin7/content/IntegratedGeodesyAndGIS
[21] http://www.mathworks.com/
[22] http://www.spss.com/
[23] http://www.terraseer.com/products_stis.php
References (II)
[24] http://www.esds.ac.uk/international/support/user_guides/gisoverview.asp
[25] http://www.ccg.leeds.ac.uk/software/gam/
[26] http://geodacenter.asu.edu/software
[27] http://www.unine.ch/CSCF/grasp/
[28] http://grass.fbk.eu/
[29] http://www.landserf.org
[30] http://www.riks.nl/products/Map_Comparison_Kit
[31] http://mapserver.gis.umn.edu/
[32] http://postgis.refractions.net/
[33] http://cran.r-project.org/web/views/Spatial.html
[34] http://regionalanalysislab.org/index.php/Main/STARS
[35] P. Compieta, S. Di Martino, M. Bertolotto, F. Ferrucci, and T. Kechadi. 2007. Exploratory spatiotemporal data mining and visualization. J. Vis. Lang. Comput. 18, 3 (June 2007), 255-279.
[36] C. Ahlberg, C. Williamson, B. Shneiderman, Dynamic queries for information exploration: an
implementation and evaluation, in: Proceedings ACM CHI’92, ACM Press, New York, 1992, pp. 619–
626. NF
[37] M. Harrower, A.M. MacEachren, A.L. Griffin, Developing a geographic visualization tool to support
earth science learning, Cartography and Geographic Information Science 27 (4) (2000) 279–293. NF
[38] W.L. Hibbard, B.E. Paul, D.A. Santek, C.R. Dyer, A.L. Battaiola, M.-F. Voidrot-Martinez, Interactive
visualization of earth and space science computations, Computer. 27 (7) (1994) 65–72.
[39] A. Buja, J.A. McDonald, J. Michalak, W. Stuetzle, Interactive data visualization using focusing and
linking, in: Proceedings IEEE Visualization’91, IEEE Computer Society Press, Washington, 1991, pp.
156–163.
[40] D. Stojanovic, S. Djordjevic-Kajan, A. Mitrovic, Z. Stojanovic, Cartographic visualization and
animation of the dynamic geographic processes and phenomena, in: Proceedings of 19th International
Cartographic Conference, Ottawa, Canada, Vol. 1, 1999, pp. 739–746.NF
References (III)
[41] Kumar, M.; Bhatt, G.; Beeson, P.; Duffy, C. Automated Detection and Spatio-Temporal
Classification of Channel Reaches in Semi-arid Southwestern US Using ASTER. American
Geophysical Union, 2006 Joint Assembly.NF
[42] Tim E. Carpenter, Methods to investigate spatial and temporal clustering in veterinary
epidemiology, Preventive Veterinary Medicine, Volume 48, Issue 4, 29 March 2001, Pages 303320.NF
[43] Fosgate, G.T., Carpenter, T.E., Case, J.T., Chomel, B.B., 2000. Time±spatial clustering of
human cases of brucellosis: California, 1973±1992. In: Proceedings of the Ninth International
Society on Veterinary Epidemiology and Economics, Breckenridge, CO NF
[44] McKenzie, J.S., Pfeiffer, D.U., Morris, R.S., 2000. Spatial and temporal patterns of vectorborne tuberculosis infection in beef breeding cattle in New Zealand. In: Proceedings of the Ninth
International Society on Veterinary Epidemiology and Economics, Breckenridge, CO NF
[45] Chudova D, Gaffney S, Mjolsness E, Smyth P (2003) Translation-invariant mixture models for
curve clustering. In: KDD ’03: Proceedings of the ninth ACM SIGKDD international conference on
Knowledge discovery and data mining, ACM, New York, NY, USA, pp 79–88 NF
[46] Ankerst M, Breunig MM, Kriegel HP, Sander J (1999) Optics: ordering points to identify the
clustering structure. SIGMOD Rec 28(2):49–60 NF
[47] Kalnis P, Mamoulis N, Bakiras S (2005) On discovering moving clusters in spatio-temporal
data. Advances in Spatial and Temporal Databases pp 364–381
[48] Birant, D.; Kut, A.. Spatio-temporal outlier detection in large databases. 28th International
Conference on Information Technology Interfaces, 2006
[49] Jeremy Mennis, Jun Wei Liu. Mining Association Rules in Spatio-Temporal Data: An Analysis
of Urban Socioeconomic and Land Cover Change.
http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9671.2005.00202.x/abstract.
[50] Y. Huang, S. Shekhar, and H. Xiong, “Discovering colocation patterns from spatial datasets: A
general approach.,” IEEE Transactions on Knowledge and Data Engineering, vol. 16, no. 12, pp.
1472–1485, 2004
[51] Feng Qian ; Liang Yin ; Qinming He ; Jiangfeng He ;. Mining spatio-temporal co-location
patterns with weighted sliding window. IEEE International Conference on Intelligent Computing
and Intelligent Systems, 2009. ICIS 2009.
[52] Results from a normative dependency analysis of geographic information standards Original Research
Article
Computer Standards & Interfaces, Volume 33, Issue 5, September 2011, Pages 485-493
Serena Coetzee
[53] Mobiiscape: Middleware support for scalable mobility pattern monitoring of moving objects in a largescale city Original Research Article
Journal of Systems and Software, Volume 84, Issue 11, November 2011, Pages 1852-1870
Byoungjip Kim, SangJeong Lee, Youngki Lee, Inseok Hwang, Yunseok Rhee, Junehwa Song
[54] Spatio-temporal assessment of food safety risks in Canadian food distribution systems using
GIS Original Research Article
Spatial and Spatio-temporal Epidemiology, In Press, Corrected Proof, Available online 14 February 2012
Leila Hashemi Beni, Sébastien Villeneuve, Denyse I. LeBlanc, Kevin Côté, Aamir Fazil, Ainsley Otten, Robin
McKellar, Pascal Delaquis
[55] Journal of Visual Languages and Computing & Computing (2007) 255–279Exploratory spatiotemporaldata mining and visualizationP. Compietaa,b, S. Di Martinoc, M. Bertolottoa, F. Ferruccic, T.
Kechadia, 5Interactive Visualization of Earth and Space Science Computations William L. Hibbard1&2,
Brian E. Paul1, Andre' L. Battaiola1&3, David A. Santek1, Marie-Francoise Voidrot-Martinez4 and Charles
R. Dyer2
[56] Tim E. Carpenter ,Methods to investigate spatial and temporal clustering in veterinary epidemiology,
Preventive Veterinary Medicine 48 (2001) 303±320
[57] Panos Kalnis1, Nikos Mamoulis2, and Spiridon Bakiras, On Discovering Moving Clusters in Spatiotemporal Data,
[58]Derya Birant, Alp Kut, Spatio-Temporal Outlier Detection in Large Databases
[59] Jeremy Mennis, Jun Wei Liu, Mining Association Rules in Spatio- Temporal Data: An Analysis of
Urban Socioeconomic and Land Cover Change, Transactions in GIS, 2005, 9(1): 5–17
[60] Yan Huang, Member, IEEE, Shashi Shekhar, Fellow, IEEE, and Hui Xiong, Discovering Co-location
Patterns from Spatial,Datasets: A General Approach
[61] Feng Qian Liang Yin Qinming He Jiangfeng He, Mining Spatio-Temporal Co-location Patterns with
Weighted Sliding Window
[62] Feng Qian Liang Yin Qinming He Jiangfeng He, Mining Spatio-Temporal Co-location Patterns
with Weighted Sliding Window
[63] Leila Hashemi Beni, Spatio-temporal assessment of food safety risks in Canadian food
distribution systems using GIS, Spatial and Spatio-temporal Epidemiology
[64] Sajimon Abraham, Spatio-temporal similarity of network-constrained moving object
trajectories using sequence alignment of travel locations, Transportation Research Part C 23 (2012)
109–123
[65] Bart Kuijpers a, Trajectory databases: Data models, uncertainty and complete query
Languages, Journal of Computer and System Sciences 76 (2010) 538–560
[66] Mauricio Marin , A meta-index for querying distributed moving object database servers ,
Information Systems 35 (2010) 637–661
[67] Jun Feng*, Modified Histogram: A Spatio-temporal Aggregate Index for
Moving Objects in Road Networks, Procedia Engineering 29(2012) 4135–4139
[68] Osman Abul , Anonymization of moving objects databases by clustering and perturbation ,
Information Systems 35 (2010) 884–910
[69] Lien-Fa Lin, Yuan-Ko Huang, Scalable processing of continuous K-nearest neighbor queries
with
uncertain velocity, Expert Systems with Applications 38 (2011) 9256–9265
[70] Agustina Buccella , Building a global normalized ontology for integrating geographic data
sources , Computers & Geosciences 37 (2011) 893–916
[71] T.E. Carpenter, The spatial epidemiologic (r)evolution: A look back in time and forward
to the future, Spatial and Spatio-temporal Epidemiology 2 (2011) 119–124
[72] Steffen Gebhardt , Improving data management and dissemination in web based information
systems by semantic enrichment of descriptive data aspects , Computers & Geosciences 36 (2010)
1362–1373
[73] Ping Fan, Vague continuous K-nearest neighborqueriesovermovingobjects with
uncertainvelocityinroadnetworks , Information Systems 37 (2012) 13–32
[74] Jaegeol Yim, A Kalman filter updating method for the indoor moving object database, Expert Systems with
Applications 38 (2011) 15075–15083
[75] HUANG Yong-qi1, Research on Development of Agricultural Geographic Information Ontology, Journal of
Integrative Agriculture
2012, 11(5): 865-877
[76] Amin Yousefi-Sahzabi, GIS modeling of CO2 emission sources and storage possibilities, Energy Procedia 4
(2011) 2831–2838
[77] Chen Zuana, The Design of Typhoon Meteorological Information System
and its Implementation Based on WebGIS, Procedia Environmental Sciences 10 ( 2011 ) 420 – 426
[78] Monia Santini, A multi-component GIS framework for desertification risk assessment
by an integrated index, Applied Geography 30 (2010) 394–415
[79] Ainhoa González, Applying geographic information systems to support strategic environmental
assessment: Opportunities and limitations in the context of Irish land-use plans, Environmental Impact Assessment
Review 31 (2011) 368–381
[80] Thomas Kistemann, Mapping water and health: current applications and future
Developments,
[81] Ken Kaneiwa, A rough set approach to multiple dataset analysis, Applied Soft Computing 11 (2011) 2538–2547
[82] Hatem A. Rashwan and Domenec Puig, Improving the Robustness of Variational Optical Flow
Through Tensor Voting,
[83] H. Yahia , Motion analysis in oceanographic satellite images using multiscale methods and the energy cascade ,
Pattern Recognition 43 (2010) 3591–3604
[84] Wikipedia.com
[85] Oracle.com
[86] Tobler W., (1970) "A computer movie simulating urban growth in the Detroit region". Economic Geography,
46(2): 234-240.
[87] Abraham T, Roddick JF. Survey of Spatio-Temporal Databases, GeoInformatica(Springer) 1999; 3 (1): 61-99.
[88] Adam NR, Janeja VP, Atluri V. Neighbourhood Based Detection ofAnomalies in High Dimension SpatioTemporal Sensor Datasets, ACMSymposium on Applied Computing, NicosiaCyprus; 2004. p. 576-583.
[89] maps.google.com