System and method for propagating classification decisions
A system and method for propagating classification decisions is provided. Text marked within one or more unclassified documents that is determined to be responsive to a predetermined issue is received from a user. The unclassified documents are selected from a corpus. A search query is generated from the responsive text. Same result documents are identified by applying inclusive search parameters to the query, applying the search query to the corpus, and identifying the documents that satisfy the query. Similar result documents are identified by adjusting a breadth of the query by applying less inclusive search parameters and identifying documents from the corpus that satisfy the query. A responsive classification code is automatically assigned to each same result document for classification as responsive documents. The similar documents are provided to the user. A responsive classification decision is received form the user for classification as the responsive documents.
1. A system for propagating classification decisions, comprising:
a corpus of unclassified documents;
a sample of the unclassified documents selected for review by a user based on a probability that each unclassified document includes responsive text, a probability that the unclassified document fails to include responsive text, and a level of precision;
a sample module to provide the sample of unclassified documents to the user;
a classification module to classify one or more of the unclassified documents in the sample by assigning a classification code to each of the one or more unclassified documents in the sample based on instructions from the user;
an identification module to identify those unclassified documents in the corpus that are similar to the classified documents;
an assignment module to assign the same classification code of the classified documents to each of the similar unclassified documents;
a validation module to perform validation of the similar unclassified documents, comprising:
a sample selection module to select a further sample of unclassified documents from the corpus of unclassified documents;
a size determination module to determine a size of the further sample based on a confidence level and upper bound of unclassified documents remaining in the corpus;
a responsive determination module to determine that one or more of the unclassified documents in the further sample have responsive subject matter; and
a review module to perform further review of the unclassified documents in the corpus when at least one of the unclassified documents in the sample has responsive subject matter; and
a processor to execute the modules.
2. A system according to claim 1 , further comprising:
a similarity module to assign a similarity value to each of the similar unclassified documents; and
a ranking module to rank the similar unclassified documents by the assigned similarity values.
3. A system according to claim 1 , further comprising:
a sample selection module to randomly select the sample of unclassified documents for review by the user.
4. A system according to claim 1 , further comprising:
an instruction module to receive instructions from the user to assign the same classification code of the classified documents to each of the similar unclassified documents.
5. A system according to claim 1 , further comprising:
a database to store at least one of a date and time of which the similar unclassified documents are identified; and
a status module to determine whether the similar unclassified documents identified are current based on the date or time.
6. A system according to claim 1 , further comprising:
an expectation module to determine a number of expected similar unclassified documents.
7. A system according to claim 6 , further comprising:
a validation module to perform validation when the number of similar unclassified documents reaches the expected number.
8. A system according to claim 1 , wherein the size determination module increases a number of the unclassified documents in the further sample by increasing the confidence level.
9. A method for propagating classification decisions, comprising:
obtaining a corpus of unclassified documents;
selecting a sample of the unclassified documents for review by a user based on a probability that each unclassified document includes responsive text, a probability that the unclassified document fails to include responsive text, and a level of precision;
providing the sample of unclassified documents to the user;
classifying one or more of the unclassified documents in the sample by assigning a classification code to each of the one or more unclassified documents in the sample based on instructions from the user;
identifying those unclassified documents in the corpus that are similar to the classified documents;
assigning the same classification code of the classified documents to each of the similar unclassified documents; and
performing validation of the similar unclassified documents, comprising:
selecting a further sample of unclassified documents from the corpus of unclassified documents;
determining a size of the further sample based on a confidence level and upper bound of unclassified documents remaining in the corpus;
determining that one or more of the unclassified documents in the further sample have responsive subject matter; and
performing further review of the unclassified documents in the corpus when at least one of the unclassified documents in the sample has responsive subject matter.
10. A method according to claim 9 , further comprising:
assigning a similarity value to each of the similar unclassified documents; and
ranking the similar unclassified documents by the assigned similarity values.
11. A method according to claim 9 , further comprising:
randomly selecting the sample of unclassified documents for review by the user.
12. A method according to claim 9 , further comprising:
receiving instructions from the user to assign the same classification code of the classified documents to each of the similar unclassified documents.
13. A method according to claim 9 , further comprising:
storing at least one of a date and time of which the similar unclassified documents are identified; and
determining whether the similar unclassified documents identified are current based on the date or time.
14. A method according to claim 9 , further comprising:
determining a number of expected similar unclassified documents.
15. A method according to claim 14 , further comprising:
performing validation when the number of similar unclassified documents reaches the expected number.
16. A method according to claim 9 , further comprising:
increasing a number of the unclassified documents in the further sample by increasing the confidence level.