Agent based e-commerce systems that react to buyers’ feedbacks – A fuzzy approach Original Research Article

In this paper, we have introduced an agent based e-commerce system which recommends products to buyers as per their preferences. Initially, the agent collects the buyers’ preferences in fuzzy or linguistically defined terms and based on this, presents them an ordered set of products. After obtaining the buyers’ feedbacks when they actually come across the products, the seller’s agent interacts with the buyer (buyer’s agent), revises the products preferential order and recommends either the same set of products or a new set of similar products with the revised preferential order. The buyer’s revised preferences are taken here as his/her feedback after he/she comes across with the actual products (presented products). Concepts of fuzzy logic and Fuzzy Linear Programming are used here to identify the buyer’s feedbacks on the initial presentation of the products. Our methodology also measures the degree of customers’ focus on the products which are finally recommended by the e-commerce agent. The product ranking obtained through buyers’ initial preferences is considered here as his/her subjective information and the available information from the agents’ presented products are taken as the objective information.