Research On Negotiation Strategy And Process Management In Cloud Service Trade

Cloud computing has been more and more popular in recent years, and its ideas are getting accepted. A lot of service centers virtualize resources into services and gather them to build service platforms. And the concept of cloud manufacturing came into being. The much number of services satisfies the different needs of service demanders, but brings huge challenges to the service distribution. Therefore, the service trade is an important research field.Because of the different requirements of service demanders and their own desires to sell services, service providers need all kinds of modes to promote service selling. Therefore, the building of all kinds of trading modes is the basic needs of service platforms. The current trading modes posted price, negotiation, auction and so on. Among them posted price and auction have got comparatively mature research results.Moreover, the manufacturing is in a trend of being more and more elaborated and professional. The idea of customization is becoming accepted by more and more people, and requests of many service demanders need to be depicted in many attributes. The establishment of service manufacturing chains brings a lot of contradiction to services traders in many issues. Therefore, in an opening and dynamic service platform, the negotiation mechanism is the best choice to solve these trade problems. And in the current big data days, the automated negotiation has been one of the important research objects in many areas.Different from single attribute negotiations which can only achieve’win-lose’ results, multi-attribute negotiations can make both parties win, called win-win. This is mainly because of the existence of negotiators’different preferences. Win-win is useful for the establishment of trade contracts, and the long term cooperation and benefits among the traders. So win-win is our goal here, and to reach it, negotiators need to know something about the preferences each other.This paper presents a bilateral service negotiation model which can help agents to achieve win-win agreements without disclosing their preference information to others. Bayesian technique is used to learn preferences of others and help the offer proposing strategy to generate a win-win offer on Pareto-efficient frontier which can facilitate the agreements and the long-term customer relationship and profitability. Two improvements are also made to improve the effectiveness of preference learning, one of which is the establishment of two-way hypothesis spaces for effectiveness when wrong behavior of others occurs because of wrong hypothesis results, and the other one is the calculation method of conditional probability in hypothesis updating by comparing sequences of issues which can enhance the accuracy of learning without having to simulating or learning the conceding strategy of others. Experimental evaluation shows that the negotiation results of our model are close to or on the Pareto-efficient frontier even though one agent establishes wrong hypotheses, which means our model can remedy this situation and finally achieve win-win agreements.Although a lot of research has been conducted in service trading areas, the real application of trading systems is developing slowly for the lack of negotiation process managements. It also leads to the single negotiation environment. In our system, we adopt a process management based on multi-mode by which the system can handle the trading process of a service itself. And in this context web services and BPEL are used to build the mode processes to make service deployments and control simple.