Online auction is popular due to the flexibility and convenience it offers to consumers. Variety of items is put up for sale due to the high demands from users in the e-market. Each item has a reserve price which is the minimum price seller is willing to accept and agree to sell. The reserve price plays an important role in determining how much profit sellers stand to gain. A low price setting will result in a sale but with less profit, whilst a high price setting may not result in a sale. Also, sellers may initiate the auction at different starting bid price. In this work, we develop a software seller agent that generates the item's reserve price using a heuristic technique by exploiting the past bidding history. The reserve price is determined based on several factors comprising of the number of competitors, the number of bidders, the duration for the auction and the degree of profit that the seller desires when auctioning. Our previous work revealed that the proposed seller strategy works well in diverse context of auction environments when agent knows the actual number of sellers and buyers. In real auction, this information is not known since auction is entirely dynamic and unpredictable. This paper investigates the performance of our seller strategy when the information about the environment is unknown or incomplete. The performance of the intelligent seller agent will be compared to different risk type agents that draw the price based on its behaviours.
展开▼
