Machine learning (ML) and advanced statistics are important tools for drawing insights from large datasets. However, these techniques often require human intervention to steer computation towards meaningful results. In this project, we build a new system for interactive analytics through pen and touch called Vizdom. Vizdom’s frontend allows users to visually compose complex workflows of ML and statistics operators on an interactive whiteboard, and the back end leverages recent advances in workflow compilation tech niques to run these computations at interactive speeds. Additionally, we are exploring approximation techniques for quickly visualizing partial results that incrementally refine over time. Different from existing approx. query processing techniques, VizDom takes the perception of the user into account to avoid unnecessary computation where the results are not perceivable by the user.
- Muhammad El-Hindi, Zheguang Zhao, Carsten Binnig and Tim Kraska. VisTrees: Fast Indexes for Interactive Data Exploration. Research Paper, HILDA 2016 (SIGMOD 2016)
- Andrew Crotty, Alexander Galakatos, Emanuel Zgraggen, Carsten Binnig and Tim Kraska. The Case for Interactive Data Exploration Accelerators (IDEAs). Research Paper, HILDA 2016 (SIGMOD 2016)
- Andrew Crotty, Alex Galakatos, Kayhan Dursun, Tim Kraska, Carsten Binnig, Ugur Çetintemel, Stan Zdonik: An Architecture for Compiling UDF-centric Workflows. PVLDB 8(12): 1466-1477 (2015)
- Andrew Crotty, Alex Galakatos, Emanuel Zgraggen, Carsten Binnig, Tim Kraska: Vizdom: Interactive Analytics through Pen and Touch. PVLDB 8(12): 2024-2035 (2015) (Best Demo)