IntelliShopper: Intelligent agents for e-commerce


PI: Filippo Menczer

Collaborators (past and present)


In this project we consider a customer's shopping experience on e-commerce sites. In this environment, shoppers must sift through the information at the e-commerce site on their own. This is a difficult task because the type and amount information provided on e-commerce sites and how the information is presented differs from company to company. In addition whenever the product information is updated -- more products are available, or prices have been reduced -- the customer goes unaware unless he/she revisits the site at a later time.

The goal of the project is to create a software agent to aid a customer who is shopping for a product on an e-commerce site. E-commerce sites allow customers to identify a product by browsing an online catalog of products or by performing a search for the product. The shopping agent would take both of these activities into consideration while at the same time giving the customer some control.

The agent will:

The information gathered autonomously by the agent is analyzed and compared with the information with which a customer was presented during his/her original visit of the e-commerce site. If the analysis results in a higher probability of the customer making the purchase then the shopping agent will alert the customer. Otherwise, the agent continues to monitor the information at the e-commerce site until the product becomes a more appealing purchase for the customer.