基于Kemeny社会选择理论的在线服务评价研究与实现

发布时间：2018-07-07 11:05

本文选题：在线服务 + 社会选择理论　；参考：《昆明理工大学》2017年硕士论文

【摘要】：近年来,随着在线服务技术的逐渐成熟与完善,网络空间中上的在线服务数量急剧增多,完成相同或相似功能的在线服务也大量增加。用户面对服务优劣选择时,用户不可能将每个在线服务都互相比较后再产生交互;还有某些不法服务提供商提供虚假的服务信息和一些恶意用户提供了不真实的评价等,导致网络空间中的在线服务的质量参差不齐,使得用户难以从大量的服务中方便快捷的选择优质的在线服务。另一方面,用户在与在线服务产生交互的过程中,由于其消费背景、消费心理、消费爱好等因素的影响,使得用户主观偏好不一致,即用户对服务的评价准则和尺度不一致,甚至可能出现矛盾和冲突。然而,传统的在线服务评价方法并未考虑用户评价标准不一致的问题,均是同等地看待所有用户的评价,而且无法抵制恶意评价的攻击,因此,常常给用户带来误导性的服务选择决策。上述问题导致在线服务之间无法比较或者说不具备公平的可比较性,所以迫切需要一种客观公正的服务优劣排序方法可有效的辅助用户进行在线服务的优劣选择决策。针对用户评价标准不一致和偏好不一致导致网络空间中的在线服务之间不具备公正的可比较性,从而用户难以选择到满意的在线服务的问题,论文提出了基于社会选择理论计算在线服务优劣的排序方法。首先,根据用户给出的用户-服务评价矩阵构建群体偏好矩阵;然后,基于群体偏好矩阵和Kemeny社会选择函数构建0-1整数规划模型;最后,通过求解该模型可得到服务的最优排序结果。该方法聚合个体偏好为群体偏好,决策符合群体大多数人的偏好且与个体偏好保持最大的一致性。通过理论分析和实验验证了该方法的合理性和有效性。实验表明,该方法能有效地解决在线服务之间的不可比较性问题,实现在线服务的优劣排序,并可以有效抵制推荐攻击,具有较强的抗操纵性。此外,本文还根据提出的方法和模型,设计并实现了在线服务优劣评价系统。
[Abstract]:In recent years, with the maturity and perfection of online service technology, the number of online services in cyberspace has increased dramatically, and the number of online services with the same or similar functions has also increased. When the user is faced with the choice of service advantages and disadvantages, it is impossible for the user to compare each online service with each other and then interact with each other; there are also some illegal service providers that provide false service information and some malicious users provide untrue evaluation, etc. As a result, the quality of online services in cyberspace is uneven, which makes it difficult for users to choose high-quality online services conveniently and quickly from a large number of services. On the other hand, in the process of interaction with online services, the subjective preferences of users are inconsistent due to the influence of their consumption background, consumer psychology, consumer preferences, and so on, that is, the criteria and scales of users' evaluation of services are not consistent. There may even be contradictions and conflicts. However, the traditional online service evaluation methods do not take into account the inconsistency of user evaluation criteria, and treat all users' evaluation equally, and can not resist the attack of malicious evaluation. Often brings to the user the misleading service choice decision. The above problems lead to the lack of comparison or fair comparability among online services, so it is urgent to use an objective and fair ranking method to assist users to make decisions on the advantages and disadvantages of online services. In view of the inconsistency of user evaluation criteria and inconsistency of preferences, there is no fair comparability between online services in cyberspace, which makes it difficult for users to choose satisfactory online services. In this paper, a ranking method is proposed to calculate the advantages and disadvantages of online services based on social selection theory. Firstly, the group preference matrix is constructed according to the user-service evaluation matrix given by the user; then, the 0-1 integer programming model is constructed based on the group preference matrix and Kemeny social selection function. By solving the model, the optimal ranking results of the service can be obtained. This method aggregates individual preferences as group preferences, and makes decisions consistent with the preferences of the majority of individuals and maintaining the greatest consistency with individual preferences. The rationality and validity of the method are verified by theoretical analysis and experiments. Experiments show that this method can effectively solve the problem of incomparability between online services, realize the ranking of the advantages and disadvantages of online services, resist recommendation attacks effectively, and have strong resistance to manipulation. In addition, according to the proposed method and model, the online service evaluation system is designed and implemented.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52

【参考文献】