Web-Based Interactions Support for Information Systems

Abstract

Introduction Work organization, business innovation and information technology increase the heterogeneity of the components of the enterprise information system. Therefore, the information system is not monolithic. Instead, it is made up of subsystems running on distributed heterogeneous IT platforms with varying conceptual and technical pieces. That is, the subsystems, by their conceptual, organizational and technical specificity, have different representations, schemas, views and implementations of the business objects and processes. Moreover, most of these subsystems are developed and implemented on case-by-case basis to satisfy particular goals (e.g., personal information system, office automation system, group work information systems, etc.), and not as well-designed elements of an enterprise information system. Pieces of data and processes are, therefore, overlapping and replicated, leading to the dual risks of (i) inconsistency of business objects schemas and views, and (ii) inefficiency of processes. Therefore, artifacts are required: (i) to allow global, coherent and transparent views of business objects and efficient distribution of processes, and (ii) to avoid, design of the IS each time new IT is introduced. This problem has been investigated under different aspects: schema integration of heterogeneous and federated databases (Batini et al., 1986) and (Pitoura et al., 1995), (Konopnicki and Shemueli, 19998), (Castano et al., 2001), information integration (Arens et al., 1993),) and (Calvanese et al.,1998), (Kwan and Fong, 1999), cooperation and coordination paradigm such as in Computer-Supported Cooperative Work (CSCW) which tends to motive and valid GroupWare used to support and enhance activities in which more than one person are involved (Schmidt, 94), (Mills, 1999), GroupWare , (Greif, 1988), (Ellis and Wainer, 1999), Workflow Systems (Casati et al., 1999), cooperative information systems (Papazoglou and Schlageter, 1997, or as a purely technical problem which may be solved by existing technologies such as Intranet/ Internet Computing, Client/Sever, Middleware (e.g., COBRA, DCOM, RMI, ODBC, JDBC and JSP/Servlets) and Web applications (Umar, 1997), Lewandowski,1998), (Fraternali, 1999) and (Crestani, 1999). We frame this problem as an interactions problem, that is, a lack of representation and implementation of interactions among subsystems of the information system with each other and with external sources (representing business partners). In fact, the community of information systems researchers has not focused on interactions as a component of the information systems as well as it has on the components of data and operations. Instead, it is considered as technical communication problem so that communication means and tools are implemented in case-by-case basis when required, that is, when breakdowns are observed. We argue that interaction is an important aspect of the information system as well as business objects and processes. In fact, interactions allow global, unified and consistent view of business objects and synergy of processes, but also make explicit emergent knowledge, which has considerable added value for the organization. Yet, interactions do not exist naturally. It is necessary to make them explicit and visible all more so since the exploding popularity of the world wide web (web) allows building distributed interacting subsystems; and accessing a variety multimedia structured, semi structured or unstructured information representing business objects and processes via web browsers. Therefore, interactions based upon the web must be analyzed, designed and implemented as a separate, orthogonal perspective that co-exists with the two existing perspectives of business objects (data) and processes. Accordingly, this paper presents a framework for: (i) Representing interactions as a separate perspective, and (ii) Constructing artifacts of new types called Interactions Support Systems. …