Recommendation Technologies for IoT Edge Devices
Download as PPTX, PDF0 likes212 views
Seda Polat Erdeniz presents an overview of recommendation approaches and how they could be applied in the AGILE Project. The document discusses content-based, collaborative filtering, and knowledge-based recommendation technologies. It also outlines four recommendation scenarios for the AGILE Project involving recommending apps, workflows/nodes, devices, and cloud services based on a user's gateway profile and similar other gateway profiles. The goal is to increase flexibility for IoT infrastructures and application development using recommendation technologies. Future work includes optimizing the recommender engine to operate locally on gateways and offline.
Recommendation Technologies for IoT Edge Devices
- 1. Seda Polat Erdeniz Institute for Software Technology 1 Seda Polat Erdeniz PhD Student / Project Assistant Institute for Software Technology (IST) Graz University of Technology, Austria
- 2. Seda Polat Erdeniz Institute for Software Technology 2 Agenda • Introduction • Recommendation Technologies • AGILE Project • Recommendation Scenarios in AGILE • Conclusions & Future Work
- 3. Seda Polat Erdeniz Institute for Software Technology 3 Introduction
- 4. Seda Polat Erdeniz Institute for Software Technology 4 Introduction The major goal of this paper is to provide an overview of recommendation approaches and to discuss their relevance for the AGILE Project.
- 5. Seda Polat Erdeniz Institute for Software Technology 5 Recommendation Technologies
- 6. Seda Polat Erdeniz Institute for Software Technology 6 Recommendation Technologies Content- Based Show me more of the same what I have liked! Collaborative Filtering Tell me what is popular among my peers! Knowledge - based Tell me what fits based on my needs Utility - based Show me the best items based on my most Important utilities
- 7. Seda Polat Erdeniz Institute for Software Technology 7 AGILE Project
- 8. Seda Polat Erdeniz Institute for Software Technology 8 AGILE Project • Supported by Horizon2020 (2016-2018) • 16 partners among Europe • Open calls for SMEs who will use AGILE gateway in their own IoT project http://agile-iot.eu/
- 9. Seda Polat Erdeniz Institute for Software Technology 9 Pilots of AGILE 5 Pilot Projects 1- Quantified Self 2- Open Field and Cattle Monitoring 3- Pollution Monitoring 4- Enhanced Retail Services 5- Port Area Monitoring http://agile-iot.eu/
- 10. Seda Polat Erdeniz Institute for Software Technology 10 Recommendation Scenarios in AGILE
- 11. Seda Polat Erdeniz Institute for Software Technology 11 Scenario-1: App Recommendation Recommendation of gateway apps to the users based on their gateway profiles; - plugged-in devices - installed apps - installed - workflows/nodes and - other similar gateways profiles
- 12. Seda Polat Erdeniz Institute for Software Technology 12 Content-Based App Recommendation
- 13. Seda Polat Erdeniz Institute for Software Technology 13 Scenario-2: Workflow / Node Recommendation Recommendation of workflows/nodes to the users based on their gateway profiles; - plugged-in devices - installed apps - installed - workflows/nodes and - other similar gateways profiles
- 14. Seda Polat Erdeniz Institute for Software Technology 14 Collaborative Filtering for Workflow/Node Recommendation
- 15. Seda Polat Erdeniz Institute for Software Technology 15 Scenario-3: Device Recommendation Recommendation of devices to the users based on their gateway profiles; - plugged-in devices - installed apps - installed - workflows/nodes and - other similar gateways profiles
- 16. Seda Polat Erdeniz Institute for Software Technology 16 Scenario-4: Cloud Services Recommendation Recommendation of cloud services to the users based on their gateway profiles; - plugged-in devices - installed apps - installed - workflows/nodes and - other similar gateways profiles
- 17. Seda Polat Erdeniz Institute for Software Technology 17 Conlusions & Future Work
- 18. Seda Polat Erdeniz Institute for Software Technology 18 Conclusion We showed how to apply recommendation technologies to increase the flexibility of IoT infrastructures in the context of application development and deployment
- 19. Seda Polat Erdeniz Institute for Software Technology 19 Future Work - Optimize the recommender engine to be able to run totally on the gateway and also offline. • Current Recommender Services works in 2 sides: • on the gateway locally as a gateway service • on the server side as a backend for the gateway service
- 20. Seda Polat Erdeniz Institute for Software Technology 20 Thanks for your attention! Any questions?
Editor's Notes
- #20: Is an AGILE project for adaptive IoT gateways which provides HW and SW solutions. 16 european partners (companies and Universities)