System and method for providing visual suggestions for document classification via injection

1. A system for providing visual suggestions for document classification via injection, comprising:

a database to store clusters of unclassified documents and a set of reference documents, each reference document associated with a classification code; and

a server comprising a central processing unit, memory, an input port to receive the clusters from the database, and an output port, wherein the central processing unit is configured to:

compare one or more of the unclassified documents within one such cluster to the reference documents;

identify for the cluster, one or more of the reference documents that are similar to the compared unclassified documents;

inject the similar reference documents into the cluster;

set a threshold for a number of the similar reference documents to be injected into the cluster; and

display each of the similar reference documents in the cluster with a visual indicator representative of the associated classification code, display the unclassified documents of the cluster, and provide a suggestion for classification for one of the unclassified documents within each cluster based on the visual indicators of the similar reference documents.

2. The system according to claim 1 , further comprising:

a reference set generator to generate the reference set, comprising:

a corpus of uncoded documents;

a similarity determination module to determine a similarity between each of the uncoded documents in the corpus;

a similarity identification module to identify those uncoded documents that are dissimilar to the other uncoded documents in the corpus;

a classification assignment module to assign the classification codes to the dissimilar uncoded documents; and

a document grouping module to group the dissimilar uncoded documents with the assigned classification codes as the reference set.

3. The system according to claim 1 , further comprising:

a reference set generator to generate the reference set, comprising:

a cluster module to cluster a set of uncoded documents;

a criteria application module to apply selection criteria to the clusters;

a document selection module to select one or more of the clusters that satisfy the selection criteria;

a classification assignment module to assign the classification codes to the uncoded documents in the selected clusters; and

a document grouping module to group the uncoded documents with the assigned classification codes as the reference set.

4. The system according to claim 1 , further comprising:

a reference set generator to generate the reference set, comprising:

a clustering module to cluster a set of uncoded documents;

a criteria application module to apply selection criteria to the uncoded documents within the clusters;

a document selection module to select a plurality of the uncoded documents from the clusters that satisfy the selection criteria;

a classification assignment module to assign the classification codes to the selected uncoded documents; and

a document grouping module to group the selected uncoded documents with the assigned classification codes as the reference set.

5. The system according to claim 1 , further comprising:

a reference set selection module to select the set of reference documents from a prior document review project.

6. The system according to claim 1 , further comprising:

a comparison determination module to determine the one or more unclassified documents for comparison with the reference documents, comprising one of:

a cluster center module to determine a cluster center based on a subset of the unclassified documents and to designate the subset of the unclassified documents in the cluster center as the one or more unclassified documents for comparison;

a sample module to generate a sample of the cluster based on a subset of the unclassified documents and to designate the subset of the unclassified documents in the sample as the one or more unclassified documents for comparison; and

a center sample module to designate the subset of unclassified documents in the cluster center and the subset of unclassified documents in the sample as the one or more unclassified documents for comparison.

7. The system according to claim 1 , further comprising:

a receipt module to receive a classification code for the unclassified document and to add the unclassified document and classification code to the set of reference documents.

8. The system according to claim 1 , wherein the visual indicators each comprise one of a color and a symbol.

9. The system according to claim 1 , wherein the one or more reference documents injected into the cluster are at least one of the same as or different from reference documents injected into other clusters.

10. A method for providing visual suggestions for document classification via injection, comprising:

obtaining clusters of unclassified documents;

obtaining a set of reference documents, each reference document associated with a classification code;

comparing one or more of the unclassified documents within one such cluster to the reference documents;

identifying for the cluster, one or more of the reference documents that are similar to the compared unclassified documents;

injecting the similar reference documents into the cluster;

setting a threshold for a number of the similar reference documents to be injected into the cluster;

displaying each of the similar reference documents in the cluster with a visual indicator representative of the associated classification code and further displaying the unclassified documents of the cluster; and

providing a suggestion for classification for one of the unclassified documents within each cluster based on the visual indicators of the similar reference documents.

11. The method according to claim 10 , further comprising:

generating the reference set, comprising:

obtaining a corpus of uncoded documents;

determining a similarity between each of the uncoded documents in the corpus;

identifying those uncoded documents that are dissimilar to the other uncoded documents in the corpus;

assigning the classification codes to the dissimilar uncoded documents; and

grouping the dissimilar uncoded documents with the assigned classification codes as the reference set.

12. The method according to claim 10 , further comprising:

generating the reference set, comprising:

clustering a set of uncoded documents;

applying selection criteria to the clusters;

selecting one or more of the clusters that satisfy the selection criteria;

assigning the classification codes to the uncoded documents in the selected clusters; and

grouping the uncoded documents with the assigned classification codes as the reference set.

13. The method according to claim 10 , further comprising:

generating the reference set, comprising:

clustering a set of uncoded documents;

applying selection criteria to the uncoded documents within the clusters;

selecting a plurality of the uncoded documents from the clusters that satisfy the selection criteria;

assigning the classification codes to the selected uncoded documents; and

grouping the selected uncoded documents with the assigned classification codes as the reference set.

14. The method according to claim 10 , further comprising:

selecting the set of reference documents from a prior document review project.

15. The method according to claim 10 , further comprising:

determining the one or more unclassified documents for comparison with the reference documents, comprising one of:

determining a cluster center based on a subset of the unclassified documents and designating the subset of the unclassified documents in the cluster center as the one or more unclassified documents for comparison;

generating a sample of the cluster based on a subset of the unclassified documents and designating the subset of the unclassified documents in the sample as the one or more unclassified documents for comparison; and

designating the subset of unclassified documents in the cluster center and the subset of unclassified documents in the sample as the one or more unclassified documents for comparison.

16. The method according to claim 10 , further comprising:

receiving a classification code for the unclassified document; and

adding the unclassified document and classification code to the set of reference documents.

17. The method according to claim 10 , wherein the visual indicators each comprise one of a color and a symbol.

18. The method according to claim 10 , wherein the one or more reference documents injected into the cluster are at least one of the same as or different from reference documents injected into other clusters.