Use Case: Common Public Service Model
During the last decade, numerous efforts have taken place for developing eGovernment applications and information systems. An important portion of such systems aim at supporting, and better organizing the electronic public service provision (e-services). All these systems create and make available to their constituents public services. These services are documented by developers in an impromptu and ad hoc way. This results in the lack of a common understanding or even definition of the “public service” concept. Each system uses its own representation and as a result they produce fragmented pieces of information with limited added value outside it own “world” as it becomes difficult to link together, reuse and combine services/information provided by one system with those provided by others. Due to this situation, it is very hard to aggregate information from various sources or combine existing services to provide a new level of service. Moreover, it is not possible to create machine-readable public service descriptions that could enable functionalities like automated service discovery and composition. Millions of web pages exist with public sector information but need a human reader to understand, process, abstract and use all this wealth that is available over the web.
The target population as potential beneficiaries includes all the stakeholders involved in the service provision process, i.e., the citizens and the businesses who consume public services and the governmental officials and industrial partners who define, develop and provide these services.
The above presented problem can be addressed by developing and agreeing on a common public service model as a technology independent generic representation of the public administration service. This public service model describes the basic concepts of a public service in a similar way that e.g. FOAF identifies and describes the basic concepts of a person’s profile or/and the SIOC ontology describes the basic concepts of on-line communities. A model like this could become an internal and core component of national/federal eGovernment interoperability strategies to ensure broad adoption and conformance. Such a model aims at:
- Creating a minimum common language for describing public services, contributes towards the homogenization of public services descriptions and facilitates the sharing and reuse of these descriptions. Governmental portals will have services described in a common way, thus enabling interoperability and cross-portal querying. The model facilitates the discovery of public services which are available online supporting complex queries and advanced search options.
- Making the service provision process identifiable, understandable and comparable to both constituents and service owners streamlining in this way process improvement and reengineering. The model is expected to help identify problems and bottlenecks in the process, e.g. by visualizing common attributes of services and making feasible simple and structured visualizations of complex services.
- Services and information annotated with concepts from the public service model could become machine-understandable. Queries like “give me all public services available for blind people” or “give me all services that are affected by a change in law x” could become possible in such environment.
Existing CASE tools could be extended by adding the concepts of the model among the standard modeling components. This would support the efficient modeling of public services and would enhance the development of eGovernment information systems and applications. The model could be formally described as an XML/RDF document or an ontology, so that it can be used for semantically annotating electronic public services. This annotation would be useful in different technology advancement stage, i.e. it could be used for annotating html descriptions of services (e.g. using RDFa or GRDDL) or could be taken one step further and combined with formal Semantic Web Service frameworks (e.g. SA-WSDL). There are already existing prototypes of the model being used with RDFa/SA-REST and WSMO based annotations in order to enrich semantic service descriptions with domain specific semantics. In addition it could also be implemented in eForms software as well as content management, query and analytical software.
Identified problems or limitations
The primary problem is to persuade system architects and/or central eGovernment units about the need and benefits of following a structured, standardized way of modelling public services and the benefits of following such practices. Thus, the most important issue is adoption and community building around the model. Experience has shown that heavy service models are difficult to be adopted by both the industry and governments (e.g. OWL-S, WSMO). The idea is to exploit the bottom-up approach for uptake followed by many successful web specs/standards (e.g. FOAF, SIOC, see next section too).
Another issue that has to be verified is to ensure that the model is generic enough so as to capture different types of public services possibly in different administrative environments and at the same time detailed enough to provide useful semantics for the annotated public services. The model should be extensible to allow domain specific specializations/extensions. The minimum common core will ensure a minimum level of interoperability while the domain specific extensions will allow the model to cover more detailed semantics depending on the particularities of each use case. From the deployment point of view, the approach seems to fit well with RDFa, Microformats and SA-REST based annotations.
There are many relevant initiatives. Some indicative follow: From eGovernment:
- The UK Common Government Information Model is a high-level information model for all activities undertaken by the public sector. It is part of the more general UK e-Service Development Framework. The model hasn’t been used and the reasons for this failure deserve further attention.
- The Business Reference Model of the Federal Enterprise Architecture in the US provides an integrated view of the Federal Government’s business, detailing activities that agencies perform to achieve their mission and function. The extend of adoption and usage of this model by US agencies is interesting to be further investigated.
- The Governmental Markup Language (GovML) proposes a metadata element set for governmental resources and a template for describing life events and public services. It is a research stimulated initiative with application in several projects and currently used in a Cypriot project for public service documentation.
- The recent CEN/ISSS eGov-Share group draft "Reference Ontology and Metadata Schema".
- Interestingly, governments (at least in EU) already discuss and include in their eGovernment interoperability strategies similar approaches.
From service literature/science:
- The OASIS SOA service model
- All Semantic Web Service literature is based on the assumption of creating and formalizing a common service model (e.g. OWL-S, WSMO, SA-WSDL, SA-REST).
From common vocabularies
- FOAF, SKOS and SIOC could be also perceived as relevant initiatives. These are community efforts (bottom-up) to create vocabularies for modeling particular domains in a generic way. Simplicity and extensibility are two core ideas behind the success of such initiatives: a) specifying a core, and simple vocabulary gives better chances for adoption while b) leaving the specs open to further extensions/specializations allows reusability and tailoring to specific needs in particular, countries, domains, and use cases.
- More generally, developing vocabularies and markup languages for specific domains has recently become a very active area.
The mission of the eGov IG is "...to explore how to improve access to government through better use of the Web and achieve better government transparency using open Web standards at any government level (local, state, national and multi-national)". The proposed use cases have been organized into three categories:
- Provide: public services on the web, either transactional or information services
- Engage: with citizens and businesses, on government terms or on the citizens terms
- Enable: public sector information re-use work
The proposed model can provide the means for describing Public Sector information using a common vocabulary, thus contributing both to the "provide" and "enable" category. The effort is also consistent with the workgroup activity entitled "Seamless Integration of Data". This implies that a standarided way of modeling public sector data is expected to decrease the interoperability burden among eGovernment systems, enabling seamless integration. The proposed model can be used with existing technologies e.g. RDF, RDFa, Microformats for annotating public information.