一种基于代理的模糊多目标决策方法

在电子商务过程中，买方用户面对的往往是多家卖方提供的类似服务．于是就出现了买方用户必须在多个卖方提出的方案中做出决策，选出提供方案最优惠的卖家，接受它的方案或者进行更深入的谈判．特别是对于基于代理的谈判系统，买方代理利用这种方法从网上无数的卖家中初步选出最合适的卖方代理显得尤为重要．对于选取合适的卖方代理，本文的处理方式是：买方向中间人代理提出自己的需求后，中间人代理通过某种方式(比如招标或者搜索)首先收集到可能满足需求的所有卖方代理，将它们的方案形式化后，在模糊机理论和TOPSIS的基础上，提出了一种定性判断和定量分析相结合的模糊多目标群决策方法，最后给出了算例。

A Fuzzy Multi-objective Decision-Making Method for Agent-based Negotiation

In the E-commerce world, the buyers always face with several sellers who provide similar services. Then, the buyers should make a decision from the solutions given by several sellers, select the best seller with the best solution, and accept his solution or negotiate further. Especially for the agent-based negotiation support system, it is very important that the buyer agent can use this method to select the most appropriate seller agent initiative from a mass of ones. To select right seller agent, the process method in this paper is: After the buyer agent summit the buyer's requirement to the broker agent, the broker collect all the seller agents which probably meet the buyer's requirement by bidding or search, and formalize their solutions. Then, based on the concepts of fuzzy set theory and TOPSIS, a fuzzy multi-objective decision-making approach is presented, in which the quantitative analysis and qualitative judgment are integrated. Finally, a numerical example is given.