Information Semantic Tools for Knowledge Discovery in an Integrated Ocean Observing System

1
Information Semantic Tools for Knowledge
Discovery in an Integrated Ocean Observing System

• Roger King, Surya Durbha, Nicolas Younan,
• S. Bheemireddy, S.K. Akamanchi

ESA-EUSC 2008 Image Information Mining
pursuing automation of geospatial intelligence
for environment and security 4-6 March 2008
2
Outline

• Semantic Enrichment-Information Semantics in an
  Integrated Ocean Observing System.
• GA approach for rapid image information mining

3
Introduction

• The Integrated Ocean Observing System (IOOS) (The
  U.S. contribution to the GOOS and GEOSS)
• Buoys and other ocean platforms
• Presently national networks
• NDBC, GoMOOS, TAO, etc.
• Data discovery and conversion problems due to
• Syntactic, structural, and semantic heterogeneity
  in the datasets.
• We are developing a services-driven Sensor Web
  Enablement framework for resolving these
  heterogeneity problems.

4
Coastal Buoys

• Buoyant platforms are usually attached to a
  specific location on the bottom of the sea to
  some submerged object.
• Navigational warnings
• Monitor oceanic and near surface atmospheric
  conditions.
• Typical measurements may include
• wind speed, direction and gust, barometric
  pressure, air temperature and water temperature
  some also have solar radiation, rainfall,
  visibility, etc. Many buoys also measure wave
  parameters, either wave height or wave direction,
  or both.

A physical diagram of a buoy station (from NOAA)
5
Buoy Ontology
6
Semantic heterogeneities
Sea surface Temperature
Ocean Temperature
GCMD (Global Change Master Directory)
Wind_Speed_ve (Vector averaged wind speed)
Wind_Speed
Wind_Speed
DODS
Wind_Speed_sc (scalar averaged wind speed)
Wind Speed
Water Temperature
NDBC
7
Semantic Conflicts

• Scaling conflicts
• The data from coastal sensor systems are highly
  heterogeneous in syntax, structure and semantics.
  
• Example Coastal-Marine Automated Network
  (C-MAN) station data typically include barometric
  pressure,
• wind direction,
• speed and gust, and
• air temperature
• some C-MAN stations are designed to also measure
  sea water temperature, water level, waves,
  relative humidity, precipitation, and visibility

GoMOOS buoy data (DODS)
National Data Buoy Center (NDBC)
8
SensorML and Descriptions for Coastal Buoys

• The general models and XML encodings for sensors
  and observation processing.
• Provides a framework within which the geometric,
  dynamic, and observational characteristics of
  sensors and sensor systems can be defined.
• Provides a functional model of the sensor system,
  rather than a detailed description of its
  hardware

(Botts and Richard, 2006).
9
SensorML Descriptions for Coastal Buoys

• A buoy station belongs to a System when described
  in SensorML.
• The SensorML defines the relative positions of
  components and the communication interfaces. It
  also provides means of referencing other
  documents in the metadata section.

A logical diagram of a buoy system.
10
Coastal Sensor Web Enablement
Decision Support Tools (monitoring, control,
emergency response)
Sensor Web Enablement (SWE)
Emergencies Support and Management

• Discovery
• Access
• Tasking
• Alerts

Risk Vulnerability assessments
Web services (Sensor observation Service, OM,
etc.) Encodings based on open
standards
Flood forecasting
Storm Surge Visualization
SWE Clients
Heterogeneous Network Sources (Various
monitoring sensors)
11
Other components of SWE

• Sensor Planning Service (SPS)
• Sensor Observation Service (SOS)
• A service by which a client can obtain
  observations from one or more sensors/platforms
  (can be of mixed sensor/platform types). Clients
  can also obtain information that describes the
  associated sensors and platforms.
• Sensor Alert Service (SAS)
• define how data collection requests are
  expressed, observations retrieved, and alert or
  alarm conditions defined.
• Web Notification Service (WNS)
• A service by which a client may conduct
  asynchronous dialogues (message interchanges)
  with one or more other services. This service is
  useful when many collaborating services are
  required to satisfy a client request, and/or when
  significant delays are involved is satisfying the
  request.

12
Sensor Web Architecture
CosemWare (AJAX Client)
Sensor Web Enablement (CosemWare)
Access to Collection of Sensors
DescribeSensor
GetObservation
GetCapabilities
Sensor Planning Service (SPS)
Observations and Measurements (OM)
Sensor Alert Service (SAS)
References
ConstrainedBy
Observables Dictionary
Sensor Observation Service (SOS)
Observation XSD
Registry/Catalog (CS-W)
SensorML mapping ltSensor Groupgt ltTempSensorsgt
ltprecipitation sensorsgt lt?gt
SensorML mapping ltSensor Groupgt ltwind direction
sensorgt ltwind speed sensorgt lt?gt
SensorML mapping ltSensor Groupgt ltsalinity
sensorgt ltSoilCarbon sensorgt lt?gt
(WMS)
Definition of the geometric, dynamic, and
observational characteristics of a sensor
Current System
Buoy Sensors Gateway
13
Coastal Sensor Web Enablement for Buoys
Semantic Web Portal
Query Reasoning
Invoke Service
Semantic Enrichment (COSemware)
Return Concepts
Query Processing Service
Reasoning Service
Ontology layer
Shared Ontology (Coastal Zone)
(SOS)
Application Ontology (Meteorology)
Application Ontology (RS Imagery)
Application Ontology (Hydrology)
Registry/ Catalog
Semantic Annotation
Image information mining module
Metadata extraction concept mapping
Metadata extraction concept mapping
(WMS)
Geospatial Semantic Annotation Tool (GSAT)
(WCS)
Sensor Web Enablement (Syntactic
Standardization)
Definition of the geometric, dynamic, and
observational characteristics of a sensor
Data Access Protocol (DAP)
Current Systems (Syntactic Level)
Coastal Sensors Gateway
NDBC
DODS Server
GES DAAC (Ocean temperature, ocean pressure,
ocean winds, ocean biology etc)
Water level, temp, salinity, River gauge data
Wind speeds, peak winds, surface wind stresses,
meteorological data
14
COSEMWare-Work in progress
15
Rapid Image Information Mining for Coastal
Disaster Events
16
Rapid Image Information Mining (RIIM)

• Genetic algorithm based Wrapper approach for
• Feature selection
• Feature Generation
• Model creation and
• Performance evaluation

17
Training samples used in the study each sample
corresponds to a region in the image.
18
Features selected by GA (Only Feature Selection)
19
Accuracy, precision, recall and F-measure
obtained using only feature selection by GA
RSet of returned regions SSet of regions
relevant to the query
(proportion of relevant regions to all the
regions retrieved)
(proportion of relevant regions that are
retrieved, out of all relevant regions )
(weighted harmonic mean of precision and recall)
20
Rapid Image Information Mining
21
Rapid Image Information Mining
22
Questions?
rking_at_engr.msstate.edu