面向稀疏城市感知的群智协同数据质量管理关键技术研究

城市数据感知是智慧城市的重要基础，而稀疏城市感知问题是制约智能城市进一步发展的阻碍。本课题将研究如何在数据质量的基础上融合群智感知技术和基于设备的感知技术，从而建立新型的群智协同城市感知技术。本课题探讨的数据质量问题体现在三个方面：真值发现、时域连续型移动群智感知、和城市协同感知和计算。我们重点研究了相关数据管理技术在质量、代价和效率这三方面的作用，以及三者之间的相互关系；随之，我们综合应用机器学习、大数据管理、近似算法、sketching等并提出了一系列相关技术。我们相信，通过本课题提出的针对城市稀疏感知的关键技术的研究，将有力地支持城市感知乃至智慧城市的建设。

Urban data sensing makes the foundation for the smart city infrastructures, while the sparse urban sensing incurs data quality issues which hinder the further development of smart cities. In this work, we study the problem of data quality management taking advantages of both mobile intelligence and sensing devices for the enhancement of collaborative urban sensing. We study data quality issues of three phases, truth inference, mobile crowdsensing, and collaborative urban sensing and computing. In particular, we study data management techniques on aspects of quality, cost, and efficiency, as well as their mutual effects during the data quality management. We investigate and propose a series of relevant techniques by utilizing the power of machine learning, big data management, approximation algorithms, sketching, and so on. We believe that the key techniques proposed in this project will provide strong theoretical and technical supports for the construction of urban sensing and smart cities.