Scaling-up shopbots - A dynamic allocation-based approach

In this paper we consider the problem of eCommerce comparison shopping agents (shopbots) that are limited by capacity constraints. In light of the phenomenal increase both in demand for price comparison services over the internet and in the number of opportunities available in electronic markets, shopbots are nowadays required to improve the utilization of their finite set of querying resources. In this paper we introduce PlanBot, an innovative shopbot which uniquely integrates concepts from production management and economic search theory. PlanBot aims to maximize its efficiency by dynamically re-planning the allocation of its querying resources according to the results of formerly executed queries and new arriving requests. We detail the design principles that drive the PlanBot's operation and illustrate its improved performance (in comparison to the traditional shopbots' First-Come-First-Served (FCFS) query execution mechanisms) using a simulated environment which is based on price datasets collected over the internet. Our encouraging results suggest that the design principles we apply have the potential of being used as an infrastructure for various implementations of future comparison shopping agents.