A GENERAL CITIZEN-SOURCING FRAMEWORK FOR CITY SERVICES

With the rise of ICT in smart cities, citizens actively contribute to city governance. Citizens contribute through sensing and processing capabilities of smart phones, thus developing an innovative collaboration paradigm, called “citizen-sourcing”, which fosters public engagement and participation in a collaborative governance. However, implementing such citizen-sourcing implies a collaborative process that is more complex than traditional bureaucratic procedures, and involves a wider range of data. Indeed, it faces various challenges, which include governance process, data, and IT architecture. Governance process challenge relates to the stakeholder engagement and participation, accountability, communication and collaboration. Data challenge includes not only data openness, generalization, heterogeneity, but also, with unstructured data, data duplication and information trustworthiness. Finally, IT architecture challenge addresses scalability, business-IT alignment, and accessibility. Our main contribution is a framework, called G-CSF (General Citizen-Sourcing Framework for city services), which defines reference collaboration processes, a reference data model, and a reference architecture. G-CSF stems from a benchmark to evaluate citizen-sourcing systems in terms of process, data, and architecture. Within G-CSF, the reference collaboration process addresses four stakeholders, namely Feed Producer, Feed Coordinator, Feed Consumer, and Feed Contractor, which, respectively, represent (a) stakeholders who submit service request (e.g., citizens, customers), (b) collaboration coordinators (e.g., municipality officers), (c) city departments who answer the service request, and (d) field staff (e.g., service response teams) who deliver services. In order to integrate heterogeneous data into a unique exchangeable format, we propose a reference data model, namely General City Feed Specification (GCFS), which facilitates the service configuration and data sharing for city service providers. In order to better support a scalable business-IT alignment and efficient data integration and processing, we define a Publisher / Subscriber architecture, which supports a distributed data-driven service composition and orchestration, and provides an easy-to-configure and easy-to-deploy environment for stream data processing; Additionally, we illustrate distinctive automated services, namely automated feed detection, credibility assessment, similarity analysis, and automated feed dispatch. The key innovations of G-CSF include: a) a supervised incremental text classifier (KLD-Star) for detecting feeds in OSN (Online Social Network); b) a hybrid credibility assessment approach for assessing information credibility of feeds, which assesses credibility on user reputation, semantics, and similarity of nearby feeds. By far, G-CSF is the first citizen-sourcing solution addressing automated feed detection, credibility assessment and de-duplication. We illustrate two case studies as proof of concept, namely CITY FEED and MOBANA. CITY FEED manages city issues, and it has been deployed in Pavia, and Basiglio, a small municipality in Milan, Italy. MOBANA focuses on the public transit domain, and proves the scalability and processing efficiency of the proposed architecture.