System and method for classifying documents via propagation
A system and method for classifying documents via propagation are provided. Text highlighted by a user within at least one document is identified. The document is associated with a classification code based on the highlighted text. At least a portion of the highlighted text is compared with one or more unclassified documents. Those unclassified documents with text that is the same as the compared portion of the highlighted text in the document are identified. The classification code of the document is automatically assigned to the identified unclassified documents with text that is the same. Further, those unclassified documents with text that is similar to the compared portion of the highlighted text in the document are identified. The unclassified documents with similar text are provided to the user for review and classification.
1. A system for classifying documents via propagation, comprising:
a text identification module to identify within at least one document, text that is highlighted by a user, wherein the document is associated with a classification code based on subject matter of the highlighted text for that document in relation to a predetermined legal issue;
a comparison module to compare at least a portion of the highlighted text with one or more unclassified documents;
a same document identification module to identify those unclassified documents with text that is the same as the compared portion of the highlighted text in the document;
a classification assignment module to automatically assign the classification code of the document to the identified unclassified documents with text that is the same as the compared portion of the highlighted text in the document;
a similar identification module to identify those unclassified documents with text that is similar to the compared portion of the highlighted text in the document;
a distribution module to provide the unclassified documents with similar text to the user for review and classification; and
a processor to execute the modules.
2. A system according to claim 1 , wherein the text identification module identifies further text within the document that is highlighted, wherein the further text is associated with an issue different from the highlighted text.
3. A system according to claim 2 , further comprising:
an issue module to distinguish between the different issues by marking the highlighted text and the further highlighted text with a different indicator comprising at least one of color, font style, or size.
4. A system according to claim 1 , further comprising:
a determination module to determine the unclassified documents with the same text comprising:
a query module to generate a search string comprising one or more words from the highlighted text;
a search application module to apply the search string to the unclassified documents; and
a document selection module to select the unclassified documents with the same text when the text is highly inclusive of the words in the search string.
5. A system according to claim 4 , further comprising:
a further document selection module to select the unclassified documents with similar text when the text is less inclusive of the words in the search string than the text that is the same as the compared portion of the highlighted text in the document.
6. A system according to claim 5 , further comprising:
a review module to perform quality control review of the search string via at least one of sensitivity analysis and sampling.
7. A system according to claim 5 , further comprising:
a delivery module to provide the search string to the user for review; and
a receipt module to receive from the user, instructions comprising one of exporting the search string, deactivating the search string, and editing the search string.
8. A system according to claim 7 , further comprising:
a deactivation module to unclassify all documents related to the search string when the search string is deactivated.
9. A system according to claim 7 , further comprising:
an edit module to receive edits to the search string;
a revision module to revise the search string; and
the deactivation module to deactivate the search string without the edits.
10. A system according to claim 1 , wherein the classification code comprises one of privileged, responsive, and non-responsive.
11. A method for classifying documents via propagation, comprising:
identifying within at least one document, text that is highlighted by a user, wherein the document is associated with a classification code based on subject matter of the highlighted text for that document in relation to a predetermined legal issue;
comparing at least a portion of the highlighted text with a plurality of unclassified documents;
identifying those unclassified documents with text that is the same as the compared portion of the highlighted text in the document;
automatically assigning the classification code of the document to the identified unclassified documents with text that is the same as the compared portion of the highlighted text in the document;
further identifying those unclassified documents with text that is similar to the compared portion of the highlighted text in the document; and
providing the unclassified documents with similar text to the user for review and classification.
12. A method according to claim 11 , further comprising:
identifying further text within the document that is highlighted, wherein the further text is associated with an issue different from the highlighted text.
13. A method according to claim 12 , further comprising:
distinguishing between the different issues by marking the highlighted text and the further highlighted text with a different indicator comprising at least one of color, font style, or size.
14. A method according to claim 11 , further comprising:
determining the unclassified documents with the same text comprising:
generating a search string comprising one or more words from the highlighted text;
applying the search string to the unclassified documents; and
selecting the unclassified documents with the same text when the text is highly inclusive of the words in the search string.
15. A method according to claim 14 , further comprising:
selecting the unclassified documents with similar text when the text is less inclusive of the words in the search string than the text that is the same as the compared portion of the highlighted text in the document.
16. A method according to claim 15 , further comprising:
performing quality control review of the search string via at least one of sensitivity analysis and sampling.
17. A method according to claim 15 , further comprising:
providing the search string to the user for review; and
receiving from the user, instructions comprising one of exporting the search string, deactivating the search string, and editing the search string.
18. A method according to claim 17 , further comprising:
unclassifying all documents related to the search string when the search string is deactivated.
19. A method according to claim 17 , further comprising:
receiving edits to the search string;
revising the search string; and
deactivating the search string without the edits.
20. A method according to claim 11 , wherein the classification code comprises one of privileged, responsive, and non-responsive.