Interactive Analysis of Graph and Time-Series Data: Enabling Technologies and Systems
MetadataShow full item record
Massive amounts of data are being generated every day. While data become more universally accessible, they are also becoming increasingly more complex. With the rise of social networks and mobile sensors, graph and time series data are gaining interest in the research community. However, because of the complexity of such data, making sense of them is still a great challenge. In this thesis, we investigate techniques and systems that enable interactive visual analysis of graph and time series data. We focus on (1) technologies that enable scalable data mining algorithms on a single machine, (2) web-based large scale visualization systems, and (3) these new tools' application on two scenarios: mobile healthcare sensor data, and I/O latency data for computer clusters. The topics of this thesis lie in the intersection of data-mining, human-computer interaction, and database systems. We believe our work will inspire more innovations for interactive interpretation of big data, and human-in-the-loop data analytics systems.