Heterogeneous Crowd-Sourced Data Analytics

Advances in computing, communication, storage, and sensing technologies have reshaped the lives of people by changing the way they live, work, interact with their environments, and even socialize. Modern information systems collect valuable information about every aspect of our lives. Such data is becoming increasingly voluminous and readily available. Data is heterogeneous, contributed by the crowd of people, coming from different sources and with diverse formats. Broadly, such data is generated mainly from three sources: Internet and Web applications, sensor networks, and mobile/wearable devices. The scale and richness of the multimodal, mixed data sources present us with an opportunity to compile the data into a comprehensive picture of individuals’ daily life facets, transform our understanding of our lives, organizations and societies, and enable completely innovative urban services, including public people and freight transportation, public safety, city resource management, environment monitoring, and social interaction assistance. However, raw data is heterogeneous, redundant, fragmented, and quality-variant, which prevents their direct use for analysis, management, forecasting and planning. Consequently, emerging data analytics targeted to their sessions, including data co-mining, data fusion, data selection, need to be studied and applied more thoroughly.