Log in ....Tribune

Monday, September 22, 2003
Feature

Relax and keep ‘Eyes’ open for best deals
Intelligent agents to make online shopping easier
Sonal Chawla

E-COMMERCE is creating a vast market place with enormous number of products and pricing options. Today, information about products and vendors is easily available on the Web with orders and payments being automated. But there are several stages in the buying process, like buying decisions, product search, purchase, payment, price discovery etc., that are time consuming. Intelligent agents when used in certain stages of these processes can increase user-productivity by carrying out certain programmable routine tasks in the background. This not only reduces the transaction costs but also improves the entire experience for the buyer.

The buying agents automatically collect information about vendors and products that meet specific needs, evaluate the offerings, take decisions about the vendors and products to investigate, negotiate the terms of the transactions, place orders and make automated payments. Agent technology has not made great progress in the need identification stage and currently agents can only help in repetitive purchases. A notification agent called ‘Eyes’ at the Amazon.com site monitors a catalogue of books and notifies customers when books of their interest are available. Once the consumer need has been identified the process enters the product brokering stage. In this stage, several agents carry out critical evaluation of the product information and market recommendations to the customers. Firefly (www.firefly.com) is one such collaboration agent that makes recommendations based on group evaluation of products. The agent compiles the evaluation as well as profile of the evaluating users thereby recommending a product based on the preferences of people with similar profiles.

Today several sites require their customers to manage their negotiation strategies on their own. It is here that the agents play a vital role in automating the process of negotiation. Kasbah (http://kasbah.media.mit.edu), for instance, is an electronic marketplace where agent programs carry out transaction with each other on behalf of the customer. These agents are proactive and once launched try to sell the goods in the market place by contacting other buying agents and negotiating the best deal with them.

Bargainfinder (http://bf.cstar.ac.com/bf) is another experimental virtual shopping agent for the Web. It queries the price and availability of user specified music CDs. AuctionBOT is a general purpose Internet auction server at the University of Michigan. Here a seller can create new auctions to sell products by choosing from a selection of auction types and then specifying its parameters. AuctionBOT differs from most other auction sites on the fact that it provides an API for users to create their own software agents to automatically compete in the AuctionBOT marketplace. Such an API provides a semantically sound interface with the marketplace. T`EAte-`E0-t`EAte is another agent that uses negotiation approach to retail sales. This agent follows the argumentative style to negotiate multiple transaction terms, for example warranties, delivery times, service contracts, return policies, loan options, gift services and other merchant value-added services.