Adaptive and transparent entity screening
Systems and methods for adaptive and transparent entity screening are provided. In an aspect, a first entity list comprising a plurality of first entity records is accessed. The first entity records comprise a plurality of first entity identifying attributes. A second entity list comprising a plurality of second entity records is also accessed, wherein the second entity records comprise a plurality of second entity identifying attributes. In certain aspects, an entity screening model pairs the first entity identifying attributes of the first entity record with the second entity identifying attributes of the second entity record, executes different matching algorithms on pairs of entity identifying attributes, and determines an overall likelihood that the pair of entity records are a match based on aggregation of match scores from the plurality of matching algorithms.
1. A computing system comprising:
one or more computer processors; and
a tangible computer readable medium storing software instructions executable by the one or more computer processors to cause the computing system to:
access a first entity list comprising a plurality of first entity records, each of the first entity records comprising a plurality of first attributes;
access a second entity list comprising a second plurality of second entity records, each of the second entity records comprising a plurality of second attributes;
determine pairs of entities each comprising a first entity of the first entity records and one of the second entities of the second entity records; and
for each determined pair of entities:
apply one or more matching rules to attributes of the pair of entities to determine a plurality of attribute specific match scores associated with the determined pair of entities; wherein each attribute specific match score is associated with a specific attribute of the first entity and a corresponding specific attribute of the second entity; and
generate an overall pair match score for each of the determined pairs of entities by combining the plurality of attribute specific match scores for the determined pair of entities, wherein the overall pair match score represents a strength of a match between the first and the second entity.
2. The computing system of claim 1 , wherein the software instructions are further configured to cause the computing system to:
provide a user interface including at least some of the pairs of entities based on overall pair match scores.
3. The computing system of claim 1 , wherein the software instructions are further configured to cause the computing system to:
periodically automatically determine changes in overall pair match scores.
4. The computing system of claim 3 , wherein the periodically automatically determining changes in overall pair match scores comprises:
accessing the first and second entity lists, including any updates to the first and second plurality of entities;
determining updated pairs of entities and applying the matching rules to each of the updated pairs to generate updated overall pair match scores; and
determining changes in overall pair match scores compared to previously generated overall match scores of the updated pair of entities.
5. The computing system of claim 1 , wherein a first matching rule comprises a matching algorithm determining a number of deletions, insertions, or substitutions required to transform a first string to a second string and a second matching rule comprises a matching algorithm encoding homophones to match strings with similar pronunciations despite minor differences in spelling, and a third matching rule comprises an exact matching algorithm.
6. The computing system of claim 1 , wherein a first and second matching rules are each associated with a first entity identifying attribute, the first matching rule comprising a deterministic matching algorithm and the second matching rule comprising a probabilistic matching algorithm.
7. The computing system of claim 6 , wherein the attribute specific match score associated with the first matching rule is larger than the attribute specific match score associated with the second matching rule.
8. The computing system of claim 7 , wherein the second matching rule is executed only in response to an output of the first matching rule indicating that the first entity identifying attributes are not an exact match.
9. The computing system of claim 2 , wherein the software instructions are further executable to cause the computing system to:
determine one or more of the overall pair match scores that exceed an alert threshold; and
update the user interface to include an indication of the determined one or more overall pair match scores.
10. The computing system of claim 2 , wherein the user interface includes controls enabling adjustment of attribute specific match scores of respective matching rules.
11. The computing system of claim 1 , wherein the software instructions are further configured to cause the computing system to:
apply a score modifier criteria to one or more entity identifying attributes; and
if the score modifier criteria is met by the one or more entity identifying attributes, increase the overall pair match score by a score modifier.
12. The computing system of claim 11 , wherein the score modifier criteria indicates one or more particular geographic regions.
13. The computing system of claim 12 , wherein application of the score modifier is performed on each pair of entities having an overall match scores above a predetermined threshold.
14. The computing system of claim 13 , wherein the predetermined threshold is zero.
15. The computing system of claim 12 , wherein the score modifier is larger than an alert threshold.
16. The computing system of claim 2 , wherein the user interface indicates the attribute specific match scores for each of the matching rules and includes user interface controls configured to receive user input causing adjustments to the attribute specific match scores.
17. The computing system of claim 2 , wherein the user interface displays pairs of entities having highest overall pair match scores.
18. The computing system of claim 2 , wherein the user interface further displays a score modifier associated with a particular entity identifying attribute, wherein the score modifier alters the overall match score of entity pairs matching a corresponding score modifier criteria.
19. The computing system of claim 3 , wherein periodically automatically comprises in real time, once every few hours, once a day, or other time intervals.
20. A method performed by one or more computer processors by executing software instructions stored in a tangible computer readable medium, the method comprising:
accessing a first entity list comprising a plurality of first entity records, each of the first entity records comprising a plurality of first attributes;
accessing a second entity list comprising a second plurality of second entity records, each of the second entity records comprising a plurality of second attributes;
determining pairs of entities each comprising a first entity of the first entity records and one of the second entities of the second entity records; and
for each determined pair of entities:
applying one or more matching rules to attributes of the pair of entities to determine a plurality of attribute specific match scores associated with the determined pair of entities; wherein each attribute specific match score is associated with a specific attribute of the first entity and a corresponding specific attribute of the second entity; and
generating an overall pair match score for each of the determined pairs of entities by combining the plurality of attribute specific match scores for the determined pair of entities, wherein the overall pair match score represents a strength of a match between the first and the second entity.