Adopting Agent-Based Situated Decision Support Framework for Managing One-to-many Negotiations with Multiple Potential Agreements

The rise of the Internet and electronic business presents new opportunities to develop flexible and effective exchange mechanisms. The information and communication technologies allow the individuals and the organizations engage in potentially fruitful interactions and joint solution search regardless of their location and time. Rise of interest in electronic negotiation systems (ENS) is a logical consequence of the today’s business and technological trends [1]. Electronic Negotiation Systems allow, as a minimum, the participants to communicate with each other exchanging offers and messages. In addition to these basic capabilities, they can also incorporate some analytical capabilities, e.g. user’s preference elicitation. The more sophisticated systems incorporate so-called software agents (active software components) that can automate various negotiation related tasks, from information search to full automation of the negotiation process. Much effort has been spent in the design and evaluation of agent solutions to automate one-to-one negotiations. Recently, researchers have been expanding the agent-based models to address bi-lateral negotiations as well. In this work our interest is in one-tomany negotiations involving multiple potential agreements. This may involve selling products or services to customers through deal-making. The work aims at applying the framework for situated decision support developed recently to this problem. The major components of situated decision support system include sensors, effectors, manager, and active user interface. We illustrate the approach through simulations for the case used previously in our agent-assisted negotiation experiments.