Electronic commerce represents a true paradigm shift. The Internet makes it feasible to integrate business processes, use new interface modalities, and extend the reach of organizations. The result is a more global, accelerated business world that affects organizations, politics, and law. IBM Research has contributed significantly to advancing the state of the practice through creating new software and to extending understanding through simulation, analysis, economics, and theory.

Future e-Marketplaces and Advanced Business-to-Business e-Commerce

Our research provided the overall structure and much of the original software for IBM's e-marketplace platform (WebSphere® Commerce Suite Marketplace Edition) and has contributed significantly to IBM's business-to-consumer and business-to-business (B2B) software. Hence, we can explore our ideas in a realistic e-commerce environment, work with customers who have unusual requirements or original business models, and transfer successes rapidly to products.

We have projects to explore future-looking e-marketplaces that will support more complex ways to describe goods and services, to collaborate on deals and negotiate about them, and to interact with other marketplaces and automated agents. Advanced forms of auctions and trades involving complex, multi-attribute products, approximate product descriptions, and bundles will permit more precise matching of needs of the market participants and lead to more efficient economies. Trading mechanisms and individual strategies can be optimized through modeling, simulation, and mining. Flexible rule systems support rapid reactions to changing market conditions.

We are also examining ways for enterprises to collaborate to reach optimal deals through multistage negotiations, or to form virtual organizations that can jointly construct new offerings. Organizations must communicate despite disparities of software platforms and business processes, so we have designed industrial-strength integration mechanisms to help automate design, manufacturing, supply web, decision-making, and marketing processes.

We are conducting fundamental research on the role of e-marketplaces in traditional supply chain relationships and are exploring ways to manage business risks encountered in B2B marketplaces, such as demand and supply uncertainty, supply chain disruptions, and production variability.

The next major phase of e-commerce will be about making business far more dynamic and enabling organizations to negotiate and consummate deals automatically by taking advantage of new standards, protocols, and processes. We are addressing challenges of deep mechanization at many levels, from protocols and legal frameworks to new business processes and online algorithms for decision and optimization.

Business-to-Consumer e-Commerce

There are also many challenges in the business-to-consumer and other individual-oriented spaces. We are actively pursuing new techniques for personalization as well as general architectural support for customization. We have projects to improve information needed for marketing, to develop techniques for decision support, and to implement sales tools such as e-coupons.

We are exploring ways to utilize location and context information in mobile commerce to improve the customer experience and to maximize the perceived value of the effort required for such interactions. Practical constraints of portable devices and wireless networks introduce difficult problems in adapting applications, content, and the human interface.

Experimental Economics

In our Experimental Economics Laboratory, we interactively test economic models of markets, organizations, and individual behavior. For example, we simulate exchanges wherein buyers and sellers negotiate prices and earn profits that are based on their assessments of value and their skill at negotiation. These experiments allow us to evaluate alternative market trading rules and gain insight into the performance of marketplaces and their infrastructures.

Information Economies

The Internet may evolve into an information economy bustling with economically motivated software agents that automatically buy, refine, and sell information, goods, and services. This may represent the convergence of business-to-business, business-to-consumer, dynamic, and marketplace e-commerce. We are studying the implications of this new breed of economic player through analysis of agent-based market models - with a focus on automated bidding and pricing by software agents - and are designing well-behaved agent strategies and interaction protocols. We have published numerous papers about the behavior of information economies, about the occasionally counter-intuitive results of learning strategies, and on competitions between humans and software agents.

Please contact Paridhi Verma to obtain copies of the Computer Science Brochure