Systems and methods for maintaining a random sample of documents
Systems and methods related to maintaining a random sample of documents that is representative of a pool of documents are provided. As documents are ingested into the pool of documents, a random number may be assigned to the documents. The documents may then be sorted into an ordered list. As the documents in the pool are provided to a review platform for manual review, the documents may be included in a review queue based at least in part on the ordered list. As additional documents are added to the pool of documents, the new documents are interleaved into the ordered list to maintain the random and representative nature of the random sample.
1. A method for maintaining a random sample of documents, the method comprising:
ingesting, via one or more processors, a first set of documents into a pool of documents;
sorting, via the one or more processors, the first set of documents into a random order to generate an initial ordering of the pool of documents;
providing, via the one or more processors, documents within the pool of documents to a review platform based at least in part upon the initial ordering of the pool of documents;
defining, via the one or more processors, the random sample of documents to be an initial set of documents within an ordering of the pool of documents associated with labels applied via the review platform, wherein the initial set of documents changes in size as additional documents are reviewed via the review platform;
ingesting, via one or more processors, a second set of documents into the pool of documents; and
interleaving, via the one or more processors, the second set of documents into the initial ordering of documents to generate an updated ordering of the pool of documents wherein the documents within the first set of documents remain in a same sequential order with respect to other documents within the first set of documents.
2. The method of claim 1 , further comprising:
analyzing, via the one or more processors, the labels associated with the random sample of documents to generate a metric associated with the pool of documents.
3. The method of claim 2 , wherein:
the labels applied via the review platform indicate a responsiveness to an inquiry; and
the metric associated with the pool of documents is a richness of documents responsive to the inquiry.
4. The method of claim 3 , further comprising:
calculating, via the one or more processors, at least one of a precision, recall, or elusion of a classifier based upon the richness metric.
5. The method of claim 2 , wherein generating the metric comprises:
tracking, via the one or more processors, historical values for the metric as additional documents are reviewed via the review platform.
6. The method of claim 1 , wherein sorting the first set of documents comprises:
assigning, by the one or more processors, documents in the first set of documents a random number; and
sorting, by the one or more processors, the first set of documents based upon the corresponding random numbers.
7. The method of claim 6 , wherein interleaving the second set of documents comprises:
assigning, by the one or more processors, documents in the second set of documents a random number; and
interleaving, by the one or more processors, the second set of documents such that the pool of documents is sorted based upon the corresponding random numbers.
8. The method of claim 6 , further comprising:
detecting, via the one or more processors, that a document has been removed from the pool of documents; and
removing, via the one or more processors, the removed document from the ordering of the pool documents.
9. The method of claim 8 , wherein:
removing the removed document from the ordering of the pool of documents maintains a correspondence between the removed document and the assigned random number.
10. The method of claim 1 , wherein providing the documents within the pool of documents to the review platform further comprises:
providing, via the one or more processors, documents within the pool of documents to a review platform based at least in part upon a priority score assigned by a classifier.
11. A system for maintaining a random sample of documents, the system comprising:
one or more processors; and
one or more non-transitory memories storing processor-executable instructions that, when executed by the one or more processors, cause the system to:
ingest a first set of documents into a pool of documents;
sort the first set of documents into a random order to generate an initial ordering of the pool of documents;
provide documents within the pool of documents to a review platform based at least in part upon the initial ordering of the pool of documents;
define the random sample of documents to be an initial set of documents within an ordering of the pool of documents associated with labels applied via the review platform, wherein the initial set of documents changes in size as additional documents are reviewed via the review platform;
ingest a second set of documents into the pool of documents; and
interleave second set of documents into initial ordering of documents to generate an updated ordering of the pool of documents wherein the documents within the first set of documents remain in a same sequential order with respect to other documents within the first set of documents.
12. The system of claim 11 , wherein the instructions, when executed by the one or more processors, cause the system to:
analyze the labels associated with the random sample of documents to generate a metric associated with the pool of documents.
13. The system of claim 12 , wherein:
the labels applied via the review platform indicate a responsiveness to an inquiry; and
the metric associated with the pool of documents is a richness of documents responsive to the inquiry.
14. The system of claim 13 , where the instructions, when executed by the one or more processors, cause the system to:
calculate at least one of a precision, recall, or elusion of a classifier based upon the richness metric.
15. The system of claim 12 , wherein to generate the metric, the instructions, when executed by the one or more processors, cause the system to:
track historical values for the metric as additional documents are reviewed via the review platform.
16. The system of claim 11 , wherein to sort the first set of documents, the instructions, when executed by the one or more processors, cause the system to:
assign documents in the first set of documents a random number; and
sort the first set of documents based upon the corresponding random numbers.
17. The system of claim 16 , wherein to interleave the second set of documents, the instructions, when executed by the one or more processors, cause the system to:
assign documents in the second set of documents a random number; and
interleave the second set of documents such that the pool of documents is sorted based upon the corresponding random numbers.
18. The system of claim 16 , wherein the instructions, when executed by the one or more processors, cause the system to:
detect that a document has been removed from the pool of documents; and
remove the removed document from the ordering of the pool documents.
19. The system of claim 18 , wherein:
removing the removed document from the ordering of the pool of documents maintains a correspondence between the removed document and the assigned random number.
20. A non-transitory computer-readable storage medium storing processor-executable instructions, that when executed cause one or more processors to:
ingest a first set of documents into a pool of documents;
sort the first set of documents into a random order to generate an initial ordering of the pool of documents;
provide documents within the pool of documents to a review platform based at least in part upon the initial ordering of the pool of documents;
define a random sample of documents to be an initial set of documents within an ordering of the pool of documents associated with labels applied via the review platform, wherein the initial set of documents changes in size as additional documents are reviewed via the review platform;
ingest a second set of documents into the pool of documents; and
interleave second set of documents into the initial ordering of documents to generate an updated ordering of the pool of documents wherein the documents within the first set of documents remain in a same sequential order with respect to other documents within the first set of documents.