Linked Data-Driven Smart Spaces
1 like727 views
This document describes a Linked Data-driven approach for enabling interactions between smart space components and end users. It presents a reference platform architecture that uses semantic annotation, recommendation systems, and user-generated content management to retrieve and enrich information about smart space components from the Linked Data Cloud. An evaluation of the approach is implemented through a tourism use case in St. Petersburg, Russia involving recommendations of local points of interest.
Linked Data-Driven Smart Spaces
- 1. 1 Oscar Rodriguez Rocha Cristhian Figueroa Iacopo Vagliano Boris Moltchanov Linked Data-Driven Smart Spaces
- 2. 2 Outline 1. Introduction 2. Reference platform and use case 3. Evaluation 4. Conclusion
- 4. 4 Context â Information about Smart Space components in Linked Data Cloud â Increasing number of mobile devices and their capabilities â Improvements in content generation on-the-go (mobile UGC) â Critical to provide to end user right information in the right moment
- 5. 5 Aims 1. Linked Data based approach for a) Enabling interactions among Smart Space components b) Enabling interactions Smart Space component â end user c) Retrieving and enriching information about Smart Space components 2. Practical application of the approach
- 7. 7 Linked Data-Driven Recommender Systems â Suggest items to users â Linked Data-based similarity â Three types â Hierarchical â Transversal â Hybrid
- 8. 8 Recommendation Algorithm Ranked set of items (classified by categories) Item (URI)
- 9. 9 Recommendation Algorithm Category Extraction Ranked set of items (classified by categories) Item (URI)
- 10. 10 Recommendation Algorithm Category Extraction Subcategory Extraction Ranked set of items (classified by categories) Item (URI)
- 11. 11 Recommendation Algorithm Category Extraction Subcategory Extraction Item Extraction Ranked set of items (classified by categories) Item (URI)
- 12. 12 Recommendation Algorithm Category Extraction Subcategory Extraction Item Extraction Item - category Mapping Ranked set of items (classified by categories) Item (URI)
- 13. 13 Recommendation Algorithm Category Extraction Subcategory Extraction Item Extraction Item - category Mapping Transversal Similarity Calculation Ranked set of items (classified by categories) Item (URI)
- 14. 14 Recommendation Algorithm Category Extraction Subcategory Extraction Item Extraction Item - category Mapping Transversal Similarity Calculation Categorization and Ranking Ranked set of items (classified by categories) Item (URI)
- 15. 15 Use Case Peter and Paul fortress Hermitage Museum St. Isaac's Cathedral St. Petersburg Smart Space
- 16. 16 Use Case Peter and Paul fortress Hermitage Museum St. Isaac's Cathedral St. Petersburg Smart Space DBpedia information +
- 17. 17 Use Case Peter and Paul fortress Hermitage Museum St. Isaac's Cathedral St. Petersburg Smart Space
- 20. 20 Conclusion 1. Linked Data based approach to manage Smart Space components information 2. Enrich information about Smart Space components 3. eToursim use case implemented with Telecom Italia
- 22. 22 Future Works â Accuracy evaluation â Integration of FI-WARE Context Management GE (context broker) â Addition of smart space components in Linked Data cloud â Additional use cases (museum)
- 23. 23 Accuracy Evaluation â Goal â Method to use â Recommender system to use for comparison â Method â Systematic review â 69 papers included â Result â User studies most used