System and method for arranging clusters in a display by theme
View Patent ↗A system and method for arranging clusters in a display by theme. Themes are found within a plurality of clusters. There are one or more concepts per theme and one or more documents per cluster. Those clusters that meet a best fit criteria are assigned by theme into a set of spines with no more than one spine formed per theme. Those spines that are unique from the other spines are retained in the set. The spines that were retained in the set are arranged in a display.
1. A system for arranging clusters in a display by theme, comprising:
a theme generator to find themes within a plurality of clusters comprising one or more concepts per theme and one or more documents per cluster; and
a spine placer, comprising:
an assigner to assign by theme those clusters that meet a best fit criteria into a set of spines with no more than one spine formed per theme;
an identifier to retain those spines that are unique from the other spines in the set; and
an arranger to arrange the spines that were retained in the set in a display.
2. A system according to claim 1 , further comprising:
a theme identifier to identify the themes by evaluating cumulative scores for the concepts determined over all of the clusters.
3. A system according to claim 2 , further comprising:
a ranker to rank the cumulative scores for the concepts; and
an evaluator to evaluate the cumulative scores against an acceptance criteria comprising at least one of cluster membership, frequency of occurrence, and user specified.
4. A system according to claim 1 , further comprising:
a selector to select candidate spines comprising a plurality of concepts identified from all of the clusters by theme as meeting an acceptance criteria to potentially become assigned to the set of spines.
5. A system according to claim 4 , wherein the acceptance criteria is defined comprising at least one of a minimum cluster reference count, maximum cluster reference count, and user specified.
6. A system according to claim 4 , further comprising:
a matcher to match the clusters to the candidate spines based on the best fit criteria.
7. A system according to claim 6 , wherein the best fit criteria comprises an equation:
F
=
log
(
popularity
rank
2
)
×
scale
where popularity comprises a count of those clusters containing one of the concepts in at least one of the candidate spines, rank comprises a ranking of the theme of the at least one candidate spine, and scale comprises a user specified bias factor.
8. A system according to claim 1 , further comprising:
a remover to remove the spines from the set, which are at least one of insufficiently large in terms of clusters and insufficiently dissimilar to the other spines in the set.
9. A system according to claim 1 , further comprising:
a grafter to graft the spines that were not retained in the set onto the spines that were arranged in the display, which share closest similarities.
10. A system according to claim 1 , wherein the spines are placed in the display based on at least one of overlap avoidance and closest similarities to those spines already arranged in the display.
11. A method for arranging clusters in a display by theme, comprising:
finding themes within a plurality of clusters comprising one or more concepts per theme and one or more documents per cluster;
assigning by theme those clusters that meet a best fit criteria into a set of spines with no more than one spine formed per theme;
retaining those spines that are unique from the other spines in the set; and
arranging the spines that were retained in the set in a display.
12. A method according to claim 11 , further comprising:
identifying the themes by evaluating cumulative scores for the concepts determined over all of the clusters.
13. A method according to claim 12 , further comprising:
ranking the cumulative scores for the concepts; and
evaluating the cumulative scores against an acceptance criteria comprising at least one of cluster membership, frequency of occurrence, and user specified.
14. A method according to claim 11 , further comprising:
selecting candidate spines comprising a plurality of concepts identified from all of the clusters by theme as meeting an acceptance criteria to potentially become assigned to the set of spines.
15. A method according to claim 14 , further comprising:
defining the acceptance criteria comprising at least one of a minimum cluster reference count, maximum cluster reference count, and user specified.
16. A method according to claim 14 , further comprising:
matching the clusters to the candidate spines based on the best fit criteria.
17. A method according to claim 16 , wherein the best fit criteria comprises an equation:
F
=
log
(
popularity
rank
2
)
×
scale
where popularity comprises a count of those clusters containing one of the concepts in at least one of the candidate spines, rank comprises a ranking of the theme of the at least one candidate spine, and scale comprises a user specified bias factor.
18. A method according to claim 11 , further comprising:
removing the spines from the set, which are at least one of insufficiently large in terms of clusters and insufficiently dissimilar to the other spines in the set.
19. A method according to claim 11 , further comprising:
grafting the spines that were not retained in the set onto the spines that were arranged in the display, which share closest similarities.
20. A method according to claim 11 , further comprising:
placing the spines in the display based on at least one of overlap avoidance and closest similarities to those spines already arranged in the display.
21. A computer-readable storage medium holding code for performing the method according to claim 11 .
22. A structured record for arranging clusters in a display by theme, comprising:
structured data comprising:
themes from within a plurality of clusters comprising one or more concepts per theme and one or more documents per cluster; and
a set of spines comprising those clusters assigned by theme that meet a best fit criteria with no more than one spine formed per theme,
wherein those spines that are unique from the other spines are retained in the set to be arranged in a display.
23. A structured record according to claim 22 , wherein the spines are removed from the set, which are at least one of insufficiently large in terms of clusters and insufficiently dissimilar to the other spines in the set.
24. Stored data for arranging clusters in a display by theme, comprising:
a plurality of clusters comprising one or more documents each;
themes identified within the clusters and comprising one or more concepts per theme; and
a set of displayable spines comprising those clusters assigned by theme and that meet a best fit criteria with no more than one spine formed per theme, wherein those spines that are unique from the other spines in the set are retained.
25. A display object for arranging clusters by theme, comprising:
a set of displayable spines, comprising:
clusters that meet a best fit criteria assigned by themes comprising one or more concepts per theme with no more than one spine formed per theme and one or more documents per cluster,
wherein retaining those spines that are unique from the other spines are retained in the set for display.
26. A display object according to claim 25 , further comprising:
a set of grafted spines comprising those spines that were not retained in the set, which share closest similarities to the set of displayable spines.