Anusuriya Devarajuanusuriya.devaraju@uni-muenster.de
1   • Background and Motivation2   • Research Goals and Questions3   • Methods4   • Theory (SEGO)5   • Application6   • Ev...
1   • Background and Motivation2   • Research Goals and Questions3   • Methods4   • Theory (SEGO)5   • Application6   • Ev...
The Sensor Revolution                                           In 2010, US Government spent around US$500 million        ...
The Sensor Web                 5
Lots of Data, and No Information?                                    6
Sensors Tell More than They Sense!            “Sensors enables an understanding of environmental                          ...
An Example: Flood Stage and Inundation                                         8
The ChallengeHow can we infer information about geographic   occurrences from sensor observations?                        ...
Sensors and Observations ModelingFormal specifications in the Semantic Sensor Webrepresent information about sensors and o...
The Challenge                11
Terminological DisagreementsIn GI Science, while the necessity for handlingtemporal phenomena has been acknowledged forsom...
Event Specifications in the Semantic Web a. Address a specific type of occurrence b. The occurrence-of-interest is not ass...
1   • Background and Motivation2   • Research Goals and Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicati...
Research Goals       Develop an ontology to capture   1       their relations.       Exploit the ontological vocabularies ...
Ontology in a Nutshell                                     Formal                    represents                           ...
Research QuestionsRequirements gathering    What are the basic representational requirements ofgeographic occurrences in t...
1   • Background and Motivation2   • Research Goals and Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicati...
The Domain of the Ontologya. Represent information about geographic occurrences from a   sensing point of view.b. Institut...
Research Methodsa. Review existing theories on occurrence and observation to1   identify the key aspects of geographic occ...
An Overview of SEGOa. Middle-out ontology development approach (Uschold:1996).b. Competency questions (Gruninger:1994).   ...
1   • Background and Motivation2   • Research Goals and Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicati...
Key Concepts of Geographic Occurrences  EXP/HIST perspectives (Galton:2006), Stimulus-centric approach (Kuhn:2009)        ...
From Observations to Occurrences  An ongoing air                A demarcated, inferredflow process acts                   ...
SEGO       25
Theoretical Insights1 Events and processes are distinguished by means of theira.   temporal shapes and their relations to ...
1   • Background and Motivation2   • Research Goals and Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicati...
Blizzard – Why It’s a Big Deal..Figure Source : http://monroetalks.com/forum/index.php?topic=12465.0   28
Definition             29
Identifying Blizzards  Weather observations supplied                                     A method for identifying blizzard...
Blizzard Application Ontology                                31
Combining Rules and Ontologies1. Domain-Specific Rules 1   Relating an observation event to its feature-of-interest    obs...
System Implementation                       A SPARQL query example.System architecture.                                   ...
1   • Background and Motivation2   • Goals and Research Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicati...
Use Case Results VerificationStation name : Brandon, ManitobaTest Data : Hourly observations (Nov-Mac, 1958-1965); 14 bliz...
An Evaluation Against SemSOS             Competency Questions                             SemSOS           SEGOSensor and ...
Analytical Research EvaluationAn evaluation approached from the System Developmentperspective (Burstein and Gregor: 1999)....
1   • Background and Motivation2   • Research Goal and Questions3   • Scope and Methods4   • Theory (SEGO)5   • Applicatio...
ContributionsBuilding blocks for developing   Applications of rules-based    application ontologies       reasoning and ev...
What’s Next? 1            2 Represent different interpretations of theDevelop          same occurrence.test cases.        ...
Thank You       For more information, please visit:SEGO Website : http://observedchange.com/ontologies/sego/       41
42
Upcoming SlideShare
Loading in …5
×

Representing and Reasoning about Geographic Occurrences in the Sensor Web

983 views
890 views

Published on

Observations are fed into the Sensor Web through a growing number of environmental sensors, including technical and human observers. While a wealth of observations is now accessible, there is still a gap between low-level observations and the high-level descriptive information they reflect. For example, we may ask what the measurements mean when a weather buoy provides a temperature time series. The challenge is not to gather a vast number of observations, but rather to make sense of them in environmental monitoring and decision making.
In order to infer meaningful information about occurrences from observations, a description of how one gets from the former to information about the latter must be expressed. This thesis develops an ontology to formally capture the relationships between geographic occurrences and the properties observed by in situ sensors. Building upon the existing positions on experiential and historical perspectives, stimulus-centric sensing, event-process algebra and thematic roles, the ontology elucidates the key concepts associated with geographic occurrences that are particularly significant from a sensing point of view. A use case for reasoning about blizzards and their temporal parts from real time series supplied by the Environment Canada illustrates the ontological approach. This thesis evaluates its findings on the basis of a comparison with an alternative approach in the Sensor Web, a verification of the use case results using an official event report published by the weather agency and an analytical assessment approached from the system development perspective.
The theoretical contribution of the thesis lies in the development of a formal model, which constitutes common building blocks for constructing application ontologies that account for inferences of geographic events from observations. With regards to its practical contribution, the thesis has demonstrated how ontological vocabularies are exploited with reasoning mechanisms to infer information about events, and to formulate symbolic spatio-temporal queries.

Published in: Technology, Education
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
983
On SlideShare
0
From Embeds
0
Number of Embeds
4
Actions
Shares
0
Downloads
11
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide
  • Formalxplicitly represents informalknowledge of a certain domain that is implicit,The knowledge captured by ontology is agreed by a group, therefore it supports knowledge sharing between computer systems.
  • The symbol grounding problem [Harnad 1990] is the challenge of giving meaning to symbols in a system by relating them to something outside the system. The ultimate candidates for that “something outside” are physical observations.
  • Representing and Reasoning about Geographic Occurrences in the Sensor Web

    1. 1. Anusuriya Devarajuanusuriya.devaraju@uni-muenster.de
    2. 2. 1 • Background and Motivation2 • Research Goals and Questions3 • Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 2
    3. 3. 1 • Background and Motivation2 • Research Goals and Questions3 • Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 3
    4. 4. The Sensor Revolution In 2010, US Government spent around US$500 million on the maintenance and operation of environmentalCrowd sourcing for Pakistan Flood Relief monitoring technology networking…. 4
    5. 5. The Sensor Web 5
    6. 6. Lots of Data, and No Information? 6
    7. 7. Sensors Tell More than They Sense! “Sensors enables an understanding of environmental variability and change”1 Basic Assumption : Their observations reflect the influence of geographic occurrences operating in the environment.1National Science Foundation Report, 2004. 7
    8. 8. An Example: Flood Stage and Inundation 8
    9. 9. The ChallengeHow can we infer information about geographic occurrences from sensor observations? 9
    10. 10. Sensors and Observations ModelingFormal specifications in the Semantic Sensor Webrepresent information about sensors and observations,but they lack details about geographic occurrences. A Functional An Ontological Ontology of O&M Analysis of O&M (Existing work on modelling sensors and observations) 10
    11. 11. The Challenge 11
    12. 12. Terminological DisagreementsIn GI Science, while the necessity for handlingtemporal phenomena has been acknowledged forsome time now, progress has been hampered by thelack of principled ways of describing these eventsand processes...(Galton:2008) “An event is comprised of processes” (Yuan:2001) “A process is composed of events” (Lemosdias et al.:2004) “One person’s process is another’s event, and vice versa” (Worboys:2005) 12
    13. 13. Event Specifications in the Semantic Web a. Address a specific type of occurrence b. The occurrence-of-interest is not associated with sensing concepts. 13
    14. 14. 1 • Background and Motivation2 • Research Goals and Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 14
    15. 15. Research Goals Develop an ontology to capture 1 their relations. Exploit the ontological vocabularies 2 with reasoning mechanisms to make inferences about geographic events. 15
    16. 16. Ontology in a Nutshell Formal represents ExplicitInformal Shared Implicit 16
    17. 17. Research QuestionsRequirements gathering What are the basic representational requirements ofgeographic occurrences in the context of the sensing domain?Formal specification : Sensing Geographic Occurrences (SEGO) How can geographic occurrences be formally modelled with respect to properties observed by sensors?Proof-of-concept implementation How can ontologies support the reasoning about geographic occurrences from sensor observations? 17
    18. 18. 1 • Background and Motivation2 • Research Goals and Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 18
    19. 19. The Domain of the Ontologya. Represent information about geographic occurrences from a sensing point of view.b. Institutionalized events are considered as the primary mode of occurrence identification.c. Observations are produced by an in-situ sensor. 19
    20. 20. Research Methodsa. Review existing theories on occurrence and observation to1 identify the key aspects of geographic occurrences.b. Develop an ontology to represent the relations between2 geographic occurrences and observations.c. Design and implement a use case; verify the use case3 results.d. Evaluate the ontology by comparing it with an alternative4 approach in the Sensor Web.e. Evaluate the research as a whole from a System5 Development perspective. (Repeat steps 1-3 if necessary) 20
    21. 21. An Overview of SEGOa. Middle-out ontology development approach (Uschold:1996).b. Competency questions (Gruninger:1994). Top Level The Descriptive Ontology for Linguistic and Cognitive Engineering (DOLCE) Ontology Domain Level SWRL Sensing Geographic Temporal Occurrences Ontology (SEGO) Ontology Application Level Blizzard application ontology 21
    22. 22. 1 • Background and Motivation2 • Research Goals and Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 22
    23. 23. Key Concepts of Geographic Occurrences EXP/HIST perspectives (Galton:2006), Stimulus-centric approach (Kuhn:2009) temporal-sub-event-of physical-object geo-process geo-event temporally- made-of participant-in geo-stimulus An analogy between events and objects, and between processes and matter. 23
    24. 24. From Observations to Occurrences An ongoing air A demarcated, inferredflow process acts high wind event as a stimulus (windspeed ≥ 40mph) actuates actuates An anemometer as a sensor windspeed 24
    25. 25. SEGO 25
    26. 26. Theoretical Insights1 Events and processes are distinguished by means of theira. temporal shapes and their relations to a sensor.2 Their relations are modeled after the analogy between eventsb. and objects, and between processes and matter.3 Functional participatory relations - relevant for queryingc. information in the Sensor Web.4 The location of an occurrence is determined by the location ofd. its participants – this does not apply to all cases!5 A feature-of-interest is regarded as an “identifiable” real-worlde. object regarding which an observation is performed.
    27. 27. 1 • Background and Motivation2 • Research Goals and Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 27
    28. 28. Blizzard – Why It’s a Big Deal..Figure Source : http://monroetalks.com/forum/index.php?topic=12465.0 28
    29. 29. Definition 29
    30. 30. Identifying Blizzards Weather observations supplied A method for identifying blizzards by the Climate Data Online1. (Lawson:2003).1http://www.climate.weatheroffice.gc.ca/climateData/canada_e.html 30
    31. 31. Blizzard Application Ontology 31
    32. 32. Combining Rules and Ontologies1. Domain-Specific Rules 1 Relating an observation event to its feature-of-interest observation-event(?e) ⋀ observed-property(?p) ⋀ feature-of-interest(?f) ⋀ has-obs-property(?e,?p) ⋀ has-bearer(?p,?f) has-foi(?e,?f)2. Application-Specific Rules2 Identifying different types of blizzard blizzard(?b) ⋀ extreme-blowing-snow(?bs) ⋀ snow-event(?s) ⋀ temporal-sub-event-of(?bs,?b) ⋀ temporal-sub-event-of(?s,?b) traditional-blizzard(?b) 32
    33. 33. System Implementation A SPARQL query example.System architecture. A time-line view. 33
    34. 34. 1 • Background and Motivation2 • Goals and Research Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 34
    35. 35. Use Case Results VerificationStation name : Brandon, ManitobaTest Data : Hourly observations (Nov-Mac, 1958-1965); 14 blizzard events A blizzard event report published A tabular view of the inferred events. by the Environment Canada. 35
    36. 36. An Evaluation Against SemSOS Competency Questions SemSOS SEGOSensor and observationsWhat are the wind-speed values and their observed time produced by station A on YYYY-MM-DD? Identify the maximum and minimum values.Events, sensing and temporal informationWhat are the observed values associated with the blizzard detected by [station id/name]on YYYY-MM-DD?Are there any ground blizzards detected by station A between YYYY-MM-DD and YYYY-MM-DD?Interrelation between eventsHow long does the blowing snow event last during the blizzard detected at station A onYYYY-MMDD?Participating entitiesWhat are the atmospheric features involved in the snow event X? 36
    37. 37. Analytical Research EvaluationAn evaluation approached from the System Developmentperspective (Burstein and Gregor: 1999). a. Significance b. Internal and external validity c. Objectivity d. Reliability Is there theoretical and practical significance? Have rival methods been considered? Are the findings congruent with or connected to prior theory? Are the study’s methods described in detail? Are the research questions clear? 37
    38. 38. 1 • Background and Motivation2 • Research Goal and Questions3 • Scope and Methods4 • Theory (SEGO)5 • Application6 • Evaluation7 • Conclusions and Future Work 38
    39. 39. ContributionsBuilding blocks for developing Applications of rules-based application ontologies reasoning and event-based querying. A formal specification that captures the relations between geographic occurrences and observations to support inferences of the former from the latter. 39
    40. 40. What’s Next? 1 2 Represent different interpretations of theDevelop same occurrence.test cases. 3 Model causality. 4 Reasoning about events across different sensors. 5 Event-oriented querying in the Sensor Web. 40
    41. 41. Thank You For more information, please visit:SEGO Website : http://observedchange.com/ontologies/sego/ 41
    42. 42. 42

    ×