Semantics-based approach for defining complex event rules for real-time algal bloom detection

1. Semantics-based approach for defining complex event rulesfor real-time algal bloom detection Jonathan Yu, Kerry Taylor, Brad Sherman 24 August 2012 CSIRO LAND AND WATER

2. Problem: “defining complex event rules for real-time algal bloom detection” Method: “Semantics-based approach” (lingua franca) Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

3. Chaffey Dam Photo credit: Brad Sherman Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

4. Chaffey Dam affected by algal blooms • Need for understanding why algal blooms happen • Historical data analysis • Various “bloom hypotheses” • Improved process for monitoring and managing the risk of algal blooms • Exploring what is happening, what are the trends • Data sifting • Lots of effort and time spent in ‘curating’ the data – field trips, modelling, consolidating disparate datasets, bringing data up to scratch so that they are analysable Photo credit: Brad Sherman Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

5. “... defining complex event rules for real-time algal bloom detection” Sensor Network WQ Weather Flow • Technological advances allow for real-time information to be published and shared through sensor networks • Mobile technology: increased capability for in-situ transmission • Platforms: OGC’s Sensor Observation Service (SOS), Global Sensor Network (GSN) • Ability to stream data from multiple sources • Enhanced monitoring capability • perform real-time analysis • Aid decision making • formulating timely and appropriate management responses Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

6. GSN VSensors Real-time stream data processing using GSN End users Programmers Curation Coding Analysis, Monitoring, Management Sensor Network High level entry for an end user e.g. Scientists and managers Knowledge hidden behind code or implicit in people’s heads Possible barrier for reusability Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

7. Data heterogeneity, integration • Multiple datasets • Often multiple data schemas and formats • Example: The use of the observation property “Total Nitrogen” • TN • N_TOT • TOTAL_NITROGEN • Want to be able to have mechanism of translating and mapping differing fields, labels to something commonly understood • Enhance interoperability Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

8. Ontologies DomainOntology SemanticSensor Net. Ontology “Semantics-based approach” • Ontologies • Capture semantics • Lingua franca • Machine readable/processable • Vocabulary of things you care about in your data • E.g. Ability to refer to ‘Total Nitrogen’ concept, rather than N_TOT • We use ontologies for: • Providing translation between fields within sensors, datasets • Defining complex event rules • Generating code for actioning complex event rules on the sensors Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

9. Ontologies GSN DomainOntology SSN Ontology VSensors Ontology-driven complex event system 1. Composes CE Ontology-enabledUser Interface 8. Views alert 2. Submits CE definition Users Semantic Mediator 7.Updates UI withalert captures CE definition Reasoner captures alerts 3. Deploys CE to GSN as VSensor via translation 6. Matching event alert generated capture rule to Vsensor mapping capture sensor / data sources mappings 5. Sensor streams data Sensor Network Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

10. Advantages of semantics-based approach Curation Coding Analysis, Monitoring, Management Curation Coding Analysis, Monitoring, Management • Transparent and transferrable • Rules, vocabularies, mappings are captured in the ontologies • Can deploy to other systems as long as they are mapped • Traceable • Alerts can attach metadata to describe triggers: what, why, when • End users can focus on exploring real-time datasets Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

11. Current work • Prototyping proof-of-concept • Simulating events using Chaffey dam study database replays • Defined sufficient domain vocabularies to be able to specify rules such as... • Given a set of data streams which monitor a range of parameters – • Total Phosphorus (TP), • Total Nitrogen (TN), • Weather conditions, • Temperature range in the water column; • Able to define some complex high-level events (e.g. potential algal bloom), given the aggregation of low-level factors (e.g. Rise in TP, TN) Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

12. Summary and future work • Presented an approach which allows event rules to be defined by end users such as research scientists • For applications such as, but not limited to, detection of algal bloom conditions • Discussed the semantics-based approach and benefits of using ontologies • Demonstrated use of ontologies for resolving data heterogeneity, ontology-driven generation of GSN Virtual Sensors and attaching metadata to alerts • Future work • Deploying on a live sensor network – interested in participating in a trial? • Application of this approach to other water quality monitoring challenges – contaminants detection, oxygen depletion Semantics-based approach for defining complex event rules for algal bloom detection | Jonathan Yu

13. Thank you Land and Water Jonathan Yu Software Engineer t +61 3 9252 6440 e jonathan.yu@csiro.au w www.csiro.au/clw ICT Centre Kerry Taylor Principal Research Scientist t +61  2 6216 7038 e kerry.taylor@csiro.au w www.csiro.au/ict Land and Water Brad Sherman Senior Research Scientist t +61 2 6246 5579 e brad.sherman@csiro.au w www.csiro.au/clw Land and Water