Open topic with navigation
ClusterSGDataGen action allows you to generate spectrograph data from a set of snapshots that you took using the
Each spectrograph dataset takes a succession of clusters from different time periods, calculates cluster similarity measures across days, and applies a graph theoretic matching algorithm. The IDOL Category component makes calculations about the conceptual spread of a cluster and its general quality. The spectrograph uses lines to represent the size (number of documents in a cluster) and quality of a cluster. The brighter a spectrograph line is, the more documents the cluster contains, and the thicker the line is, the higher the cluster quality.
All spectrograph data sets that you generate are stored in the
Sgdata subdirectory of the
Cluster directory in your IDOL Category component installation directory.
You can set up a schedule that runs the
ClusterSGDataGen action at regular intervals.
You can retrieve the spectrograph image, data, or documents by using the
For more information about the
ClusterSGDataGen action and the results that it returns, refer to the IDOL Server Reference.