基于Agent的让步态度更新的情感劝说提议生成模型——社会学习视角

Emotional Persuasive Proposal Generation Model Based on Agents'Concession Attitude Updating:A Social Learning Perspective

人工智能与在线社交网络的快速发展极大促进了社会化商务的价值创造,针对社会化商务环境下基于Agent的情感劝说中让步影响因素量化和让步更新过程研究的不足,从社会学习视角出发,提出了一种基于让步态度更新的情感劝说提议生成模型。首先,定义让步态度,并使用多元回归分析对初始让步态度进行量化;其次,对让步态度进行标准化处理,并利用动态阈值分类法划分让步态度类型;再次,采用网络结构分析法量化情感关系亲密度,并构建基于社会学习的让步态度更新算法;最后,结合前景理论,建立基于让步态度更新的情感劝说提议生成模型。实验结果表明,所提出模型能有效提高谈判效率和谈判各方的联合效用。

The rapid development of artificial intelligence and online social networks has significantly advanced social commerce.However,research on quantifying concession factors and updating processes in agent-based emotional persuasion within social commerce is limited.This paper proposes a model for emotional persuasive proposal generation based on concession attitude updating from a social learning perspective.First,concession attitude is defined,and multiple regression analysis is used to quantify the agents initial concession attitude.Next,the initial concession attitude is standardized,and dynamic threshold classification categorizes concession attitude types.Emotional relationship intimacy is then quantified using network structure analysis,and a concession attitude updating algorithm based on agent social learning is constructed.Finally,combined with prospect theory,an emotionally persuasive proposal generation model based on the updated concession attitude is established.Experimental results demonstrate that the model effectively improves negotiation efficiency and the joint utility of all parties.