Towards Transparent E-Government Systems

Government transparency should be embedded in the designs of e-Government systems. Lack of transparency can prevent the public from participating actively in government operating, such as raising questions and protesting ill-advised decisions, which in turn may cause concealing official graft or favoritism. However, it is also difficult to guarantee that designs of e-Government systems are consistent with real requirements in terms of having desired properties, such as transparency. In this thesis, we pursue, in general, a better understanding and effective gauging method of government transparency through a specific combination of formal methods (the OTS/CafeOBJ method), software engineering and domain engineering’s views. Traditional methodologies in social sciences are normally judging the transparency and other properties of an e-government system afterward; we are proposing a method from software engineering side of view, to evaluate the transparency and other desired properties from the starting stage of a government project. To achieve above goal, we carry out our work as follows: Firstly, provides preliminary knowledge on Domain Methodology, CafeOBJ algebraic specifications, the OTS/CafeOBJ method, and its’ application example in e-Government system analysis. It focuses on introducing how to apply domain engineering methodology into real world system understanding and how the structural induction of the OTS/CafeOBJ method works for software design/specification verification of invariant properties by using the CafeOBJ system as an interactive theorem prover. Secondly, presents the formal domain description based on the domain understanding firstly, and describes a General Service Delivery (GSD) model from public administration domain; and gives the CafeOBJ specification of it as well. The following transparency property discussion is based on this model. Then, a literature overview around transparency is presented; the narrative definition of transparency is also refined. And based on the narrative description of transparency, we give the formal understanding and discussion of transparency based on the GSD model. At last, we test our method on three government application cases and propose a public government license language (PGLL), which has potential possibility to improve transparency based on the previous formal discussion, and employ the OTS/CafeOBJ method to specify the PGLL as well. Our method is not panacea for solving all government transparency problems for sure. However, we are providing a constructive suggestion from views of software engineering, formal method and domain engineering for supporting to realize a more transparent eGovernment system.