Visual Data Mining of Unevenly-Spaced Event Sequences

We present a process for the exploration and analysis of large databases of events. A typical database is characterized by the sequential actions of a number of individual entities. These entities can be compared by their similarities in sequence and changes in sequence over time. The correlation of two sequences can provide important clues as to the possibility of a connection between the responsible entities, but an analyst might not be able to specify the type of connection sought prior to examination. Our process incorporates extensive automated calculation and data mining but permits diversity of analysis by providing visualization of results at multiple levels, taking advantage of human intuition and visual processing to generate avenues of inquiry.

Alex Godwin, Remco Chang, Robert Kosara, and William Ribarsky, Visual Data Mining of Unevenly-Spaced Event Sequences, IEEE VAST Posters, 2008.

@inproceedings{Godwin:VASTPoster:2008,
	year = 2008,
	title = {Visual Data Mining of Unevenly-Spaced Event Sequences},
	author = {Alex Godwin and Remco Chang and Robert Kosara and William Ribarsky},
	booktitle = {IEEE VAST Posters},
	abstract = {We present a process for the exploration and analysis of large databases of events. A typical database is characterized by the sequential actions of a number of individual entities. These entities can be compared by their similarities in sequence and changes in sequence over time. The correlation of two sequences can provide important clues as to the possibility of a connection between the responsible entities, but an analyst might not be able to specify the type of connection sought prior to examination. Our process incorporates extensive automated calculation and data mining but permits diversity of analysis by providing visualization of results at multiple levels, taking advantage of human intuition and visual processing to generate avenues of inquiry.},
}