Detecting and describing visible features on a visualization
AssigneeInternational Business Machines Corporation
InventorEser Kandogan
Life of this patent
Estimating the term…
Abstract
Embodiments of the invention relate to detecting and describing visible features of a data set on a visualization. Visible features among a set of data in a view-space are detected. The visible features include potential data clusters and trends. These visual features are characterized using data-space. The characterized detected features are overlaid on visualization for supporting interaction and exploration of the data. Detected features are explored across two or more clusters for comparison of select data.
Claims20
A computer program product for exploration of a data set, the computer program product comprising a computer-readable storage medium having program code embodied therewith, the program code being executable by a processor to:
- detect a first cluster;
- and
- characterize the first cluster, including identifying one or more first characteristics associated with the first cluster, wherein the one or more first characteristics comprise a first trend, wherein the first trend is a visual depiction of distribution of data within the first cluster, and wherein the identification of the first trend comprises the program code to:
- recognize a first shape of the first cluster based on the first visual feature;
- associate the first shape to a trend-type;
- identify one or more dimensions exhibiting the trend-type;
- and
- determine a strength of the trend-type from the identified one or more dimensions, including program code to calculate average values associated with each identified dimension and perform a standard error based on the calculated average values;
- and
- overlay the identified one or more first characteristics on a visualization for supporting interaction, the supported interaction including exploration of the identified one or more first characteristics for comparison of select data across two or more clusters.
A system comprising:
- a processing unit in communication with data storage, the data storage comprising at least one data set;
- a functional unit in communication with the processing unit, the functional unit including tools to support data exploration, the tools comprising:
- a detection manager to detect a first cluster;
- and
- a characterization manager in communication with the detection manager, the characterization manager to characterize the first cluster, including identifying one or more first characteristics associated with the first cluster, wherein the one or more first characteristics comprise a first trend, wherein the first trend is a visual depiction of distribution of data within the first cluster, and wherein the identification of the first trend comprises the characterization manager to:
- recognize a first shape of the first cluster based on the first visual feature;
- associated the first shape to a trend-type;
- identify one or more dimensions exhibiting the trend-type;
- and
- determine a strength of the trend-type from the identified one or more dimensions, including the characterization manager to calculate average values associated with each identified dimension and perform a standard error based on the calculated average values;
- a visualization manager in communication with the characterization manager, the visualization manager to overlay the identified one or more first characteristics on a visualization;
- and
- an interaction manager in communication with the visualization manager, the interaction manager to support interaction based on the overlay, the supported interaction including exploration of the identified one or more first characteristics for comparison of select data across two or more clusters.
Dependencies are inferred from the claim text.