Contextualised Cognitive Perspective for Linked Sensor Data
1 like426 views
This document proposes a contextualized cognitive approach for linking sensor data. It discusses using ontologies to describe sensor concepts and contexts. Current solutions like sensor ontologies and context classification architectures have limitations. The proposal is to use a cognitive approach inspired by human cognition, delimited to the sensor environment. Context would be described using domain-agnostic, event, and upper-level ontologies. Future work includes validating ontology choices and implementing the approach with user feedback. The goal is improving understanding of sensor reality by using the linked open data cloud to enhance classification and emulate human cognition.
![Digital Enterprise Research Institute www.deri.i
e
Self and Context awareness
Context
– Any information that can be used to characterize the situation of
entities (i.e. whether a person, place or object) that are considered
relevant to the interaction between a user and an application,
including the user and the application themselves [DeyAbowd2000]
External context
– Measured by hardware sensors
– I.e. location, light, sound, movement, touch,
Internal context
– Specified by the user or captured monitoring the user’s
interaction
– I.e. user’s goal, tasks, work context, business processes](https://image.slidesharecdn.com/ssniswc2010-101117133342-phpapp01/85/Contextualised-Cognitive-Perspective-for-Linked-Sensor-Data-3-320.jpg)
Contextualised Cognitive Perspective for Linked Sensor Data
- 1. © Copyright 2008 Digital Enterprise Research Institute. All rights reserved. Digital Enterprise Research Institute www.deri.i e Myriam Leggieri, Alexandre Passant, Manfred Hauswirth DERI NUI Galway, Ireland A contextualised cognitive perspective for Linked Sensor Data
- 2. Digital Enterprise Research Institute www.deri.i e Overview 1. Self and context awareness 1. Benefits 2. Research Question 2. Current solutions 1. Sensor ontologies 2. Context classification architectures 3. Research Challenges 3. Our proposal 1. Contextualised cognitive perspective 2. Ontology alignments and extension 4. Conlusions and Future Work
- 3. Digital Enterprise Research Institute www.deri.i e Self and Context awareness Context – Any information that can be used to characterize the situation of entities (i.e. whether a person, place or object) that are considered relevant to the interaction between a user and an application, including the user and the application themselves [DeyAbowd2000] External context – Measured by hardware sensors – I.e. location, light, sound, movement, touch, Internal context – Specified by the user or captured monitoring the user’s interaction – I.e. user’s goal, tasks, work context, business processes
- 4. Digital Enterprise Research Institute www.deri.i e Self and Context awareness : benefits Self-awareness benefits → Issue: Find all the sensors acquiring oceanographic data in Cancùn – Solution: Self-awareness: auto-determine • The kind of data acquired • The location → Issue: Observation understanding – Solution: Machine-understandable description of oservations and measurements How is the ocean like at Cancùn right now?
- 5. Digital Enterprise Research Institute www.deri.i e Context-awareness benefits Self and Context awareness : benefits Imagine: Calm ocean detected … but at the same time … another SN detects a movement of earth plates → Issue: Is it generally associated with storm surges? Solution: Search the LoD cloud
- 6. Digital Enterprise Research Institute www.deri.i e Self and Context awareness: research challenge Improvements in – Hazardous detection – Sensor retrieval – Sensor data clustering What is needed 1. Proper ontologies to support detailed descriptions 2. Effective context classification architecture
- 7. Digital Enterprise Research Institute www.deri.i e Current solutions – sensor ontologies Sensor features to be described – Sensor / Device, Capabilities, Process, Physical properties, Observation, Networks Current ontologies specialized missed in covering all those features – SWAMO - Interoperability Sensor Web products / Sensor Web services – MMI Device and CSIRO sensor ontologies - System and capabilities, Process composition, Operational and Response model With the except of W3C SSN-Xg ontology – Covers all the basic sensor features; foreseen further integrations
- 8. Digital Enterprise Research Institute www.deri.i e Current solutions – context classification architecture SOCAM (Service-oriented Context-Aware Middleware) – Centralized context interpreter COBRA (Context Broker Architecture) – Agent based; centralized context broker (KB, inference, acquisition, etc.) Context Toolkit – P2P architecture, still needs a centralized discoverer
- 9. Digital Enterprise Research Institute www.deri.i e Current solutions: research challenges Challenges: 1. Sensor ontologies: 1. Develop and choose the right ontologies to integrate with the SSN-XG one 2. Context classification 1. Storage space issue; one point of failure 2. P2P: network boundaries; user responsability 3. Data not linked together – unless because of classification output; but in this way it is not widely reusable
- 10. Digital Enterprise Research Institute www.deri.i e Our proposal: contextualized cognitive perspective 1. Contextualized cognitive approach to sensor data classification Cognitive: inspired by associative nature of human cognitiveness Contextualized: delimited to the sensor environment Human Memory LOD Cloud C Environment Unlimited, unweak, decentralized
- 11. Digital Enterprise Research Institute www.deri.i e Our proposal: ontology alignments and extension 1. Ontology support to context description Domain-agnostic ontology to describe sensor-related concepts Event modelling ontology Upper-level ontology Additional concepts: SensorProject, SensorRole, SensorHierarchy
- 12. Digital Enterprise Research Institute www.deri.i e Our proposal: ontology alignments and extension Sensor-related concept descriptions
- 13. Digital Enterprise Research Institute www.deri.i e Our proposal: ontology alignments and extension Context and situation related concept descriptions
- 14. Digital Enterprise Research Institute www.deri.i e Conclusions and future work Task: improvement of sensor reality understanding LoD cloud as – Enhancement to classification – New mean for human cognitiveness emulation Steps – Ontologies extended and aligned – Validation of ontology modelling choices – Continue with the implementation including user feedback Great Challenge: – Effectively browsing the LoD cloud