Aboozar Barzegar - Islamic Azad University, Lamerd Branch, Department of Engineering, Lamerd, Iran
Aboozar Hemmati - Islamic Azad University, Lamerd Branch, Department of Science, Lamerd, Iran

In this paper, we present a model based on weighted OCEAN model of personality for buyer and seller agents in multi agent-based electronic marketplaces. Four important factors: personality, reputation, quality, and price are considered in the model. We take into account the fact that buying agents with different personality can have differentpriorities on reputation, quality, and price of their goods and selling agents adjust their bids according to buyingagents preferences. The personality of buyer and seller agents influences their behavior in market. Seller agents adjusttheir bids according to buyer agents' preferences and their own personality. In addition, seller agents applyreinforcement learning to model the reputation of buyer agents. On the other hand, Buyer agents use their ownpersonality to assess the value of seller agents’ bids. Also buyer agents apply reinforcement learning to evaluate thereputation of seller agents based on their personality and then focus their trading on the most reputable sellers. We have implemented this model with Aglet. The results show that selling/buying agents that model the reputation of buying/selling agents based on their personality achieve more satisfaction in comparison with the selling/buying agents who only use the reinforcement learning.

E-Marketplace, Agent, Seller, Buyer, Personality