Preference-based data representation framework

1. A method of data representation, comprising:

acquiring, by a processor, preference information from a user;

generating, by the processor, rank scores of objects based at least in part on the preference information;

grouping, by the processor, the objects into one or more clusters of objects based on the rank scores;

projecting the one or more clusters of objects into one or more planes;

decomposing at least one of the one or more planes into multiple parallel planes, wherein each of the parallel planes corresponds to a different range of scores; and

generating, by the processor, a visualization of the one or more clusters of objects.

2. A method of data representation, comprising:

acquiring, by a processor, preference information from a user;

generating, by the processor, rank scores of objects based at least in part on the preference information, wherein generating the rank scores includes

determining pairwise preference degrees based on the preference information,

determining uni-criterion scores based on the pairwise preference degrees, and

determining the rank scores based on the uni-criterion scores;

grouping, by the processor, the objects into one or more clusters of objects based on the rank scores; and

generating, by the processor, a visualization of the one or more clusters of objects.

3. The method of claim 2 wherein the preference information further includes one or more comparison thresholds.

4. The method of claim 2 wherein the preference information includes search criteria information and categorical values or numerical values that indicate user-perceived relative importance of the search criteria.

5. The method of claim 4 wherein the categorical values include sub-categorical values that indicate user-perceived relative importance of the search criteria within a single category.

6. The method of claim 2 wherein the preference information includes search criteria information and pairwise comparison values, wherein a pairwise comparison value is associated with user-perceived relative importance of a pair of search criteria.

7. The method of claim 4 further comprising transforming the categorical values into the numerical values.

8. The method of claim 7 wherein transforming the categorical values into the numerical values comprises:

computing a pairwise comparison matrix based on the categorical values; and

determining the numerical values based on the pairwise comparison matrix.

9. The method of claim 5 further comprising transforming the sub-categorical values into numerical values.

10. The method of claim 2 wherein

determining the pairwise preference degrees based on the preference information comprises determining the pairwise preference degrees based on indifference and preference thresholds derived from the preference information.

11. The method of claim 2 wherein grouping the objects into the one or more clusters of objects comprises performing a K-means clustering method.

12. The method of claim 2 further comprising projecting the one or more clusters of objects into one or more planes.

13. The method of claim 12 wherein projecting the one or more clusters into the one or more planes comprises performing a radial coordinate visualization projection method.

14. The method of claim 12 further comprising decomposing at least one of the one or more planes into multiple parallel planes, wherein each of the parallel planes corresponds to a different range of scores.

15. The method of claim 14 wherein distance between the parallel planes corresponds to difference between average scores of the parallel planes.

16. The method of claim 12 further comprising optimizing visualization features of the one or more projected clusters of objects and generating the visualization based on the optimized visualization features.

17. The method of claim 16 wherein optimizing the one or more visualization features comprises generating graphical representations of the objects in the one or more planes.

18. The method of claim 17 further comprising re-positioning or re-scaling the graphical representations to minimize overlapping between the graphical representations.

19. A computer readable medium embodying a program of instructions executable by a machine to perform steps for data representation, the steps comprising:

acquiring preference information from a user;

generating rank scores of objects based at least in part on the preference information;

grouping the objects into one or more clusters of objects based on the rank scores;

projecting the one or more clusters of objects into one or more planes;

decomposing at least one of the one or more planes into multiple parallel planes, wherein each of the parallel planes corresponds to a different range of scores; and

generating a visualization of the one or more clusters of objects.

20. A system comprising:

a memory device for storing computer readable program code; and

a processor in communication with the memory device, the processor being operative with the computer readable program code to perform steps for data representation, the steps comprising

acquiring preference information from a user,

generating rank scores of objects based at least in part on the preference information,

grouping the objects into one or more clusters of objects based on the rank scores, projecting the one or more clusters of objects into one or more planes;

decomposing at least one of the one or more planes into multiple parallel planes, wherein each of the parallel planes corresponds to a different range of scores; and

generating a visualization of the one or more clusters of objects.