Database Support for Virtual Enterprises

Principal Investigator: Ahmed K. Elmagarmid

Visiting Professors: B. Benatallah, A. Bouguettaya

Visiting Researchers: Y. Saygin, Y. Villate

A virtual enterprise (VE) is an enterprise that merely exists in cyberspace. A VE's components are drawn from dedicated software that are autonomous in nature. The ultimate goal of a VE is the ability to create, manage, run, and evolve an enterprise more efficiently than its physical counterpart. Tapping into the web creates a market that does not discriminate based on an enterprise size and income. We identified the following issues to make VEs a reality:

  1. VEs need to be modeled in a systematic way. In this respect, there is a need to determine the fundamental parts of any VE as well as their attributes. Furthermore, the types of relationships that my exist at the intra and inter-VE levels need to be investigated.
  2. Models of business transactions at the intra and inter-VE levels are crucial for efficiency running an enterprise. The challenge is to come up with a model that takes full advantage of the web and internet.
  3. Internet-enabled workflows are a fundamental part of managing and running a VE. We consider the problem of dynamically expanding and integrating workflows in widely distributed, heterogeneous, and autonomous systems.
  4. Agents are a fundamental component to "glue" together the different VEs in such a way that transacting on the web becomes the only efficient and viable way of transacting as opposed to more traditional business transacting. An agent-based infrastructure has the ability to pro-actively discover and accumulate knowledge that can be used to provide intelligent, efficient, and flexible monitoring of transactions.
  5. We will investigate the use of several standards including: Web (e.g., XML), distributed object technologies (e.g., CORBA, DCOM, JavaBeans) and other technologies as enabling technologies for VEs.

1998
Annual Research Report

Department of
Computer Sciences