群智协同计算作为一种新兴计算模式能够在网络环境下进行群智服务的有效管理、实现最大化的群体智能。由于群智协同计算的开放性、多样性和复杂性，群智服务可信性面临着严峻的挑战,为保证群智系统能够高效、稳定地运行，本项目针对自适应动态的可信服务激励机制展开了深入研究，主要研究内容有：1，针对服务质量的可信性需求，提出基于真值发现的信任管理机制和基于取反方案的评价修复机制，完成局部信任聚集与全局信任优化的融合；2，针对可信行为激励需求，提出合作与信任感知的用户选择机制、自适应审计的用户行为监控机制和基于用户表现的公平支付机制，实现对群智协同过程的监控和引导；3，针对机制性能评估，提出异质用户的特征构建、用户行为的演化动力学建模和均衡状态的稳定性分析，实现对机制性能的客观预测。通过本项目研究，期望解决信任管理构建、可信服务激励，机制性能评估三个问题，加快推进群智协同计算在物联网、大数据等领域的应用。

Crowd-based collaborative computing, as a newly emerging computing paradigm, can effectively manage the crowd service and maximize the crowd intelligence in the cyberspace. Due to the openness, diversity， and complexity of crowd-based collaborative computing, the credibility of crowd services faces serious challenges. In order to guarantee the crowd system operating in an efficient and stable way, in this project, we study adaptive dynamic incentive mechanisms for the emergence of trustworthy services in crowd-based collaborative computing from three facets: (i) regarding to the demand of trustworthy services, we propose a trust management mechanism based on truth inference and a rating repairing mechanism based on the reverse method, as a result, the local aggregated trust and the global optimized trust can be well integrated;(ii) regarding to the demand of promoting the trustworthy behavior of the crowd, we propose a cooperation and trust-aware worker recommendation mechanism, a self-adaptive auditing mechanism for monitoring the crowd, and a fairness payment scheme based on workers’ performances, as a result, the process of collaborative interactions can be well monitored and guided;(iii) regarding to the demand of performance evaluation for the proposed mechanism, we propose a model for describing the heterogeneous users, study the evolutionary dynamics of users’ behaviors, and analyze the stability of each equilibrium state, as a result, the performance of the proposed mechanism can be objectively predicted. Through our research, we expect to realize the design of trust management, the encouragement of trustworthy service, and the evaluation of systems’ performances, and then promote the application of crowd-based collaborative computing into IoT (Internet of thing) and big data, etc.