System and method for placing candidate spines into a display with the aid of a digital computer

1. A method for placing candidate spines into a display with the aid of a digital computer, comprising:

generating candidate spines each comprising two or more clusters of documents, wherein each cluster is associated with one or more concept scores for concepts identified within the documents of that cluster;

generating a concept score vector for each candidate spine by aggregating as spine concept scores, the cluster concept scores of the clusters associated with that candidate spine;

normalizing each spine concept score in the spine concept score vector by dividing the spine concept score by a length of the spine concept score vector;

ordering the candidate spines by a number of the clusters associated with each candidate spine;

identifying those candidate spines sufficiently dissimilar from the other candidate spines based on the candidate spine concept score vectors;

placing the sufficiently dissimilar candidate spines in a display; and

placing at least a portion of the unplaced candidate spines in relation to the placed candidate spines.

2. A method according to claim 1 , further comprising:

discarding those candidate spines having one of concepts referenced by more than 10% of the clusters and that do not have a predetermined number of clusters.

3. A method according to claim 2 , wherein the predetermined number of clusters is five or more.

4. A method according to claim 1 , further comprising:

identifying the sufficiently dissimilar candidate spines based on a measure of similarity of each candidate spine with each other candidate spine.

5. A method according to claim 4 , further comprising:

calculating the similarity of each candidate spine to each previous candidate spine in the ordering; and

selecting the candidate spines with a low similarity to each of the other candidate spines as sufficiently dissimilar.

6. A method according to claim 1 , further comprising:

ordering the unplaced candidate spines based on the number of clusters prior to placement in the display.

7. A method according to claim 1 , further comprising:

obtaining further clusters of documents; and

assigning one or more of the clusters of documents to at least one of the placed candidate spines.

8. A method according to claim 1 , further comprising:

generating the cluster concepts, comprising:

determining a cumulative score for each concept appearing in the documents of each cluster;

ranking the concepts by the cumulative scores;

applying acceptance criteria to the ranked concepts; and

selecting those concepts that satisfy the acceptance criteria as the cluster concepts.

9. A method according to claim 1 , wherein the acceptance criteria comprise at least one of inclusion of one such concept in a center of one such cluster, inclusion of the concept in a minimum number of documents in the cluster, and inclusion of the concept in a predetermined percentage of documents in the cluster.