Summary: Voila: Visual Anomaly Detection and Monitoring with Streaming Spatiotemporal Data
Xiaofan Jiang
This paper present Voila (visual analysis of spatiotemporal data), a visual analytics system and framework for interactively detecting anomalies in spatiotemporal data collected from a dynamic streaming data source. The high-level goals of the proposed system are to process large scale, dynamic stream data to detect anomalies and allows human inspection and interpretation to guide final machine processes. In addition, it features online data processing pipeline that remains connected to data inputs and uses a tensor-based algorithm to produce descriptive patterns over time and space. By demonstrating the application of the “smart city” using the proposed system, the author demonstrates the effectiveness of Voila by case studies that quantitatively evaluate the performance.
1. What are the key points of this paper and how does it contribute to the field of visual analytics?
The objective of Voila is to process data at large scale, and dynamic to detect anomalies, also allows user to interpreted and guide the final machine process. Four features at high levels are introduced in order to achieve this objective:
• Online Data Processing Pipeline that remains connected to data inputs
• Uses a tensor-based algorithm to produce descriptive patterns over time and space
• Incorporates unsupervised Machine Learning Techniques during human interactions
• Shifts between map modes dependent on user goals

2. What are the paper’s strengths?
GIS Data in big data scenarios were used in this paper to evaluate the performance of the system. Data are transformed into a sequence of tensor time series at the granular level of an hour, a day, a week, or month.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

3. What are the papers weaknesses (what could be improved)?

4. Is this paper still relevant and will it have an impact in the future?