Visual analysis of multidimensional clusters

1. A method for visualizing a data set, comprising:

obtaining a clustering of a multidimensional dataset comprising a plurality of entities, wherein said entities are instances of a particular concept and wherein each entity comprises a plurality of features;

generating a visual icon for at least one of said entities, said visual icon having a plurality of regions, wherein each region corresponds to one of said features of said at least one entity, and wherein a size of each region is based on a value of said corresponding feature, wherein at least one data cluster is encoded into said visual icon using a hierarchical encoding scheme comprising a multi-level hierarchy of cluster as a first level object, feature type as a second level of the hierarchy, wherein said second level is below said first level, and entity as a third level of the hierarchy, wherein said third level is below said second level, to provide a visualization simultaneously having a plurality of said cluster level, said feature type level and said entity level of granularity; and

updating one or more clusters in said generated visual icon in response to one or more user interaction commands received using an input/output interface.

2. The method of claim 1 , wherein said values of said corresponding features are normalized.

3. The method of claim 1 , wherein each of said features are uniquely encoded.

4. The method of claim 3 , wherein each of said features are encoded with a unique color and wherein a distribution of said colors indicates a distribution of said corresponding feature value.

5. The method of claim 1 , further comprising the step of merging a plurality of said visual icons into a single visual icon.

6. The method of claim 5 , wherein an outer shape of said single visual icon conveys one or more cluster statistical measures.

7. The method of claim 5 , wherein said merging step further comprises the step of decomposing said plurality of visual icons into said feature values and then regrouping said feature values into said single visual icon.

8. The method of claim 5 , further comprising the step of animating transitions of said plurality of visual icons into said single visual icon following said merging step.

9. The method of claim 1 , further comprising the step of splitting at least one of said visual icons for a plurality of said entities into a plurality of visual icons.

10. The method of claim 9 , further comprising the step of animating transitions of said at least one visual icon into a plurality of visual icons following said splitting step.

11. The method of claim 1 , wherein each of said visual icons conveys one or more statistical measures.

12. The method of claim 11 , wherein an outer shape of each visual icon conveys said one or more statistical measures.

13. The method of claim 11 , wherein one or more of a color, hash pattern or shading of each of said plurality of regions conveys said one or more statistical measures.

14. The method of claim 1 , further comprising the step of employing a stabilized Voronoi-based icon layout algorithm to substantially maintain a stability of Voronoi regions when cluster changes occur.

15. The method of claim 1 , further comprising the step of employing a stabilized Voronoi-based icon layout algorithm to substantially maintain a predefined order for regions within a visual icon that places Voronoi regions next to each other according to semantic similarities.

16. The method of claim 1 , further comprising the step of embedding said visual icon for at least one of said entities in a visualization of said multidimensional dataset.

17. A system for visualizing a data set, comprising:

a memory; and

at least one processor, coupled to the memory, operative to:

obtain a clustering of a multidimensional dataset comprising a plurality of entities, wherein said entities are instances of a particular concept and wherein each entity comprises a plurality of features;

generate a visual icon for at least one of said entities, said visual icon having a plurality of regions, wherein each region corresponds to one of said features of said at least one entity, and wherein a size of each region is based on a value of said corresponding feature, wherein at least one data cluster is encoded into said visual icon using a hierarchical encoding scheme comprising a multi-level hierarchy of cluster as a first level object, feature type as a second level of the hierarchy, wherein said second level is below said first level, and entity as a third level of the hierarchy, wherein said third level is below said second level, to provide a visualization simultaneously having a plurality of said cluster level, said feature type level and said entity level of granularity; and

update one or more clusters in said generated visual icon in response to one or more user interaction commands received using an input/output interface.

18. An article of manufacture for visualizing a data set, comprising a non-transitory machine readable recordable medium containing one or more programs which when executed implement the steps of:

obtaining a clustering of a multidimensional dataset comprising a plurality of entities, wherein said entities are instances of a particular concept and wherein each entity comprises a plurality of features;

generating a visual icon for at least one of said entities, said visual icon having a plurality of regions, wherein each region corresponds to one of said features of said at least one entity, and wherein a size of each region is based on a value of said corresponding feature, wherein at least one data cluster is encoded into said visual icon using a hierarchical encoding scheme comprising a multi-level hierarchy of cluster as a first level object, feature type as a second level of the hierarchy, wherein said second level is below said first level, and entity as a third level of the hierarchy, wherein said third level is below said second level, to provide a visualization simultaneously having a plurality of said cluster level, said feature type level and said entity level of granularity; and

updating one or more clusters in said generated visual icon in response to one or more user interaction commands received using an input/output interface.