Hui Tang 

2016/07

​

The MetaCurve system aims to help analysts interpret and analyze large number of time serial measurements. Through multiple operations of aggregating periodical data, MetaCurve reveals daily patterns by showing data possibility ranges and detects anomaly by identifying and visualizing outliers.

​

The dataset of IEEE VAST 2016 Mini Challenge 2 contains more than four hundred time-serial data in two weeks of duration. Some of the data are captured by sensors monitoring building temperatures and HVAC system, while others are proximity data captures employee movements inside the building. These data are connected and affects each other. Most of the data are periodical with oneday cycle. Anomalies also existed in the data. To identify typical patterns and issues of concern from the two weeks of building and proximity sensor data, MetaCurve was used to overview, re-organize, and visualize variabilities and
anomalies for large number of periodical dataset.
 

​

The live demo website can be accessed via:

https://goo.gl/Szc1rf

​​

Role in the team:

Lead designer and developer

​

MetaCurve received the awards:

• Honorable Mention of Quality Aesthetics from 2017 IEEE Visual Analytics for Science and Technology (VAST) Challenge Mini-Challenge 2