Saturday , June 23 2018

Volume 24-Issue1-2015-VERCRUYSSEN

Collaborative Recommender System Development with
Ubiquitous Computing Capability for Risk Awareness*

Simona VARLAN2, Cristian AMANCEI3

1 ISSCO – International Software Solutions,
Bacău, Romania
2 “Vasile Alecsandri” University of Bacau,
Bacău, Romania
3 Bucharest University of Economic Studies,
Bucharest, Romania

Abstract: This paper outlines the architecture prototyping and development of ubiquitous computing for recommender systems in case of risk, with mobile and geographical capabilities based on collaborative and knowledge-based filtering approaches. Heterogeneous types of strategies developed towards minimizing risks occurrence effects are applied so as to efficiently allocate the efforts for reliable software development. A novel technique to improve risk assessment and validation at the time of event announcement involving risk type, risk severity level, users trust level, and users location is described and implemented within a dedicated mobile GIS recommender system for risk-awareness. The main purpose of research is to enhance the process of consolidation of the development cycle of GIS applications for early community risk-awareness.

Keywords: Collaborative filtering, Recommender Systems, GIS, Risk management.

<Full text
Nick VERCRUYSSEN, Cosmin TOMOZEI, Iulian FURDU, Simona VARLAN, Cristian AMANCEI, Collaborative Recommender System Development with Ubiquitous Computing Capability for Risk Awareness, Studies in Informatics and Control, ISSN 1220-1766, vol. 24 (1), pp. 91-100, 2015.

  1. Introduction

Mobile and ubiquitous computing devices are endowed with multimedia capabilities for the geospatial information processing. Information packages are sent by the members of the community to each other or to the authorities and include the geo-coordinate, with the appropriate accuracy, photos from the risk areas and data captured by means of the device sensors. To reduce the effects of the identified risk, geospatial data is processed and added on specific layers by means of classes of objects in recommendation packages for the other members of the community. Furthermore, the cross platform facility of ArcGIS Runtime SDK is implemented for Windows Phone, iOS and Android being available to all community members that actually have a smartphone or a tablet. Mapping functions and geocoding of locations is implemented as well, in order to sustain the spatial risk analysis. Taking into consideration different types of risk notification recipients in the community and the potential increased volume of information packages during alert notifications, the most relevant chronological order of priority is utilized for actual delivery of messages.

Simultaneously, the potential delay is proactively analysed, measured and optimized using initiation of extra cloud computing infrastructure in order to support the acute increase of system performance, within predefined limits.

For the optimization of user experience, a special type of interface design is required so as to overcome the restrictions generated by the size of screen. The user interface has a dual role, to assist the collaborative work for the announcement of a potential or a real observed risk and to automatically advise the community members, mainly from the risk area, in case of need. Trust and reputation is measured by weighted indicators. On the first basis, a small community is considered as target group and further development will expand on larger communities.

An imperative objective of recommender systems development is to look forward to the future generation of mobile oriented software, with a private cloud back-end approach, which has more precise and complex algorithms, large amounts of data, and large-scale data mining functionalities, so as to provide high-quality recommendations and advice to the users.

At least two main types of risks are to be managed by the recommender system, natural risks which arise from systems whose existence is beyond the human agent such as landslides, forest fires, floods, extreme weather, and risks derived from human activities, especially technological risks: pollution, severe accidental toxic emissions, explosions, fire, exposed electric wires or damaged gas pipelines. Traffic restrictions, temporarily closed roads or inadequate transport infrastructure can be easily managed and also different types of social risks such as violent and armed individuals are to be minimized. All these types of risks are first identified by common location and common time window as shared contextual information for the mobile recommender system.

Section 2 presents recent scientific achievements in the domain of mobile recommender systems MRS. A non-exhaustive list of popular mobile applications for natural and social risk management is also provided along with a brief description of each application. Section 3 describes the system requirements and architecture and section 4 presents aspects about the risk evaluation and validation of measures. Section 5 describes the ways data is processed and section 6 analyses the required infrastructure for the created mobile recommender system.

Next section tackles the integration with other data providers and last section summarizes the work.


  1. KOMPAN, M. BIELIKOVA, Personalized Recommendation for Individual Users Based on the Group Recommendation Principles, Studies in Informatics and Control, vol. 22(3), 2013, pp. 331-342.
  2. GROZAVU, A., S. PLESCAN, M. CIPRIAN MARGARINT, Indicators for the Assessment of Exposure to Geomorphologic and Hydrologic Processes, Environmental Engineering and Management Journal, vol. 12, Iss. 11, 2013, pp. 2203-2210.
  3. ADOMAVICIUS, G., A. TUZHILIN. Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions. IEEE Transaction on Knowledge and Data Engineering, vol. 17(6), 2005, pp. 734-749.
  4. KABASSI, K., Personalizing Recommendations for Tourists. Telematics and Informatics, vol. 27(1), 2010, pp. 51-66.
  5. WERTHNER, H., F. RICCI e-Commerce and Tourism. Communications of the ACM, vol. 47(12), 2004, pp. 101-5.
  6. NOGUERA, J. M., M. J. BARRANCO, R. J. SEGURA, L. MARTÍNEZ, A Mobile 3D-GIS Hybrid Recommender System for Tourism, Information Sciences, 2012, pp. 37–52.
  7. YANG, W.-S., S.-Y. HWANG, iTravel: A Recommender System in Mobile Peer-to-Peer Environment, The Journal of Systems and Software, 2013, vol. 86, pp. 12-20.
  8. BRIGUEZ, C. E., M. C. D. BUDÁN, C. A. D. DEAGUSTINI, A. G. MAGUITMAN, M. CAPOBIANCO, G. R. SIMARI, Argument-based Mixed Recommenders and Their Application to Movie Suggestion, Expert Systems with Applications, vol. 41, 2014, pp. 6467-6482.
  9. CARRER-NETO, W., M. L. HERNANDEZ-ALCARAZ, R. VALENCIA-GARCIA, F. GARCIA-SANCHEZ, Social Knowledge-based Recommender System. Application to the Movies Domain. Expert Systems with Applications, vol. 39(12), 2012, pp. 10990-11000.
  10. BARRAGANS-MARTINEZ, A. B., E. COSTA-MONTENEGRO, J. C. BURGUILLO, M. REY-LOPEZ, F. A. MIKIC-FONTE, A. PELETEIRO, A Hybrid Content-based and Item-based Collaborative Filtering Approach to Recommend TV Programs Enhanced with Singular Value Decomposition, Information Sciences, vol. 180(22), p. 2010, pp. 4290-4311.
  11. LEE, S. K., Y. H. CHO, S. H. KIM, Collaborative Filtering with Ordinal Scale-based Implicit Ratings for Mobile Music Recommendations, Information Sciences, vol. 180(11), 2010, pp. 2142-2155.
  12. TAN S., J. BU, CH. CHEN, X. HE, Using Rich Social Media Information for Music Recommendation Via Hypergraph Model, ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 7 (1), 2011, art. 7.
  13. PORCEL, C., A. TEJEDA-LORENTE, M. A. MARTINEZ, E. HERRERA-VIEDMA, A Hybrid Recommender System for the Selective Dissemination of Research Resources in a Technology Transfer Office, Information Sciences, vol. 184(1), 2012, pp. 1-19.
  14. BOBADILLA, J., F. SERRADILLA, A. HERNANDO, Collaborative Filtering Adapted to Recommender Systems of e-Learning, Knowledge Based Systems, vol. 22, 2009, pp. 261-265.
  15. SUSNEA, , G. VASILIU, D. E. MITU, Enabling Self-Organization of the Educational Content in Ad Hoc Learning Networks, Studies in Informatics and Control, vol. 22(2), 2013, pp. 143-152.
  16. SAMARAJIVA, R, N. WAIDYANATHA, Two Complementary Mobile Technologies for Disaster Warning, info, Vol. 11(2), 2009, pp. 58-65.
  18. LOUHISUO, M., Y. RAUSTE, K. ANDERSSON, T. HÄME, J. AHOLA, T. MORIHOSHI, Use of SAR Data for Natural Disaster Mitigation in the Mobile Environment. Proceedings of the 2004 Envisat & ERS Symposium, 2004.
  19. ROßNAGEL, H., J. ZIBUSCHKA, J. MUNTERMANN, T. SCHERNER, Design of a Mobile Service Platform for Public Events – Improving Visitor Satisfaction and Emergency Management, Electronic Government and Electronic Participation: Joint Proceedings of Ongoing Research and Pojects of IFIP eGOV and ePart, vol. 33, 2010, p. 193-202.
  20. VALTONEN, E., R., ADDAMS-MORING, T. VIRTANEN, A. JRVINEN, M. MORING, Emergency Announcements to Mobile User Devices in Geographically Defined Areas, Proceedings of Information Systems for Crisis Response and Management, 2004, pp. 151-156.
  21. BONTOS, D., D. VASILIU, A Pilot Web-based System for Environmental Health Impact Assessment of Air Pollution, Studies in Informatics and Control, vol. 21(2), 2012, pp. 127-136.
  23. (July, 15, 2014)
  24. BRIDGE, D., M. GOKER, L. MCGINTY, B. SMYTH, Case-based Recommender Systems, The Knowledge Engineering review, vol. 20(3), 2006, pp. 315-320.
  25. FERRETTI, V., S. POMARICO, Ecological Land Suitability Analysis through Spatial Indicators: An Application of the Analytic Network Process Technique and Ordered Weighted Average Approach, vol. 34, 2013, pp. 507-519.


* This paper is based on two previous presentations at the Second International Conference on Natural and Anthropic Risks ICNAR 2014 with the titles Requirements Analysis and User Scenario for Mobile GIS Recommender Systems Development and Colmars – Collaborative Risk Awareness Recommender System Development with Ubiquitous Computing Capability.