A Bilateral Negotiation Model Based On Well-founded Semantics

In today's society,negotiation has permeated all aspects of human activities.From bargaining between the seller and the buyer to the summit of heads of state,these are different forms of negotiation.The automatic negotiation is a hot topic in artificial intelligence.In recent years,people use non-monotonic answer set programs to express the knowledge of agents,and complete the negotiation by updating the answer set programs.Although there are a lot of methods based on answer set programs have been proposed in negotiation,but these methods have some limitations.These studies do not fully take the balance between the optimal solution and the complexity of the negotiation process into account.The answer sets are considered as the initialization requirements of the negotiation participant in study of negotiation based on the answer set programs.However,finding answer sets is a NP-complete problem.But the well-foundedmodelsof logic programs can be computed in polynomial time.Therefore,in this paper,we combine well-founded semantics with belief revision to optimize the negotiation process,and then propose a negotiation model based on well-founded semantics for bilateral negotiation.This model can effectively shorten the negotiation process,and achieve the balance between the optimal solution and the complexity of the negotiation process.In this paper,we propose a bilateral negotiation model based on well-founded semantics.Firstly,the answer set programs is used to represent all the knowledge of the negotiation participants.Then,divide the demands of both sidesby calculating the well founded model.Among that,the well-founded literal set and the other literals set are regarded as the initialization requirements,andall the literals in the unfounded literal set are not discussed.And range the literals of initial negotiation demand in descending order by the priority rules.Finally,the negotiation process is implemented by using the alternating offers protocol.Two agents gradually reduce their belief from the initial negotiation demands to update their ASPs.Thereby,an optimal result of the consultation can be obtained in acceptable time.If there are multiple negotiation results,the priority ranking rules of the negotiation results are used to sort,and the higher priority negotiation result is reserved.The main work of this paper includes the following three aspects:(1)Proposing of the priority ranking rules of literal and literal setsEach agent in negotiation has more than one requirement(also means literal)in general.A rational agent prefers to have a higher priority literal with no doubt.The priority among literals is directly related to the outcome of the negotiation.Therefore,the priority ranking rules for literals are given firstly.Based on these rules,this paper presents the priority ranking rule for literal sets.(2)Proposing of negotiation decision model based on belief revisionThe belief set for each agent can be divided into three parts: the well-founded set,the unfounded set and the other literal set.With the thought of belief revision,the three parts can be regarded as the following beliefs: the belief that negotiators preferred to be preserved,the belief that negotiatorsdo not care about;the belief that can be retained secondarily.Based on this point,this paper puts forward a negotiation decision model by the way of belief revision.(3)Proposing of the evaluation criteria for negotiation resultsIf the proposal is accepted by each participants in the process of negotiation,this proposal is a feasible trade set.After a number of rounds of negotiation between the two participants,there may be a number of feasible trade sets,that is,there are a number of feasible solutions.It is necessary to establish a reasonable and effective evaluation criteria.Therefore,this paper proposes a feasible and effective evaluation criterion of negotiation results according to the final revenue.In order to show the effectiveness and practicability of the proposed model,we apply thismodel to experiments.The experimental results show that the proposed model can solve the negotiation problems quickly and efficiently.