Custodian disambiguation and data matching
Provided is a technique for matching different user representations of a person in a plurality of computer systems may be provided. The technique includes collecting information sets about user representations from a plurality of computer systems; normalizing the information sets to a unified format; grouping the information sets in the unified format into indexing buckets based on a user name using a non-phonetic algorithm; determining a similarity score for each pair of information sets in each of the indexing buckets; classifying each information set pair into a set of classes based on the similarity scores, wherein the set of classes comprise at least matches and non-matches; and using a data structure for merging information of information set pairs classified as matches.
1. A system, the system comprising:
a processor; and
storage coupled to the processor, wherein the storage stores program instructions, and wherein the program instructions, when executed by the processor perform:
collecting information sets about user representations from a plurality of computer systems, wherein the user representations comprise heterogeneous formats, and wherein each of the user representations is used to access an application of a plurality of applications; and
matching the user representations by:
normalizing the information sets into normalized information sets, wherein each of the normalized information sets comprises one or more attributes, and wherein each of the one or more attributes has a weight;
grouping the normalized information sets into a plurality of buckets using user names;
for each pair of the normalized information sets in each of the plurality of buckets,
determining a sub-similarity score for each corresponding attribute by comparing each attribute in that pair; and
in response to determining the sub-similarity score,
determining a weighted similarity score for each corresponding attribute in that pair based on the sub-similarity score and the weight for that attribute; and
determining a similarity score for the pair by comparing each normalized information set in that pair based on each weighted similarity score for each corresponding attribute; and
merging each pair classified as a match based on the similarity score for that pair to integrate the user representations as matched user representations, wherein the merging uses a combination of a Friend-of-a-Friend ontology, a Resource Description Format, and a Web Ontology Language; and
performing a hold on data that a user has access to, wherein the user and the data are identified using the matched user representations for that user.
2. The system according to claim 1 , wherein a non-phonetic algorithm is any one of a q-gram algorithm and a suffix algorithm, and wherein the non-phonetic algorithm is used to group the normalized information sets.
3. The system according to claim 1 , wherein a longest common sub-string is determined using dynamic programming algorithms and is used to determine the similarity score.
4. A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions being executable by at least one processor of a computer to cause the computer to perform:
collecting information sets about user representations from a plurality of computer systems, wherein the user representations comprise heterogeneous formats, and wherein each of the user representations is used to access an application of a plurality of applications; and
matching the user representations by:
normalizing the information sets into normalized information sets, wherein each of the normalized information sets comprises one or more attributes, and wherein each of the one or more attributes has a weight;
grouping the normalized information sets into a plurality of buckets using user names;
for each pair of the normalized information sets in each of the plurality of buckets,
determining a sub-similarity score for each corresponding attribute by comparing each attribute in that pair; and
in response to determining the sub-similarity score,
determining a weighted similarity score for each corresponding attribute in that pair based on the sub-similarity score and the weight for that attribute; and
determining a similarity score for the pair by comparing each normalized information set in that pair based on each weighted similarity score for each corresponding attribute; and
merging each pair classified as a match based on the similarity score for that pair to integrate the user representations as matched user representations, wherein the merging uses a combination of a Friend-of-a-Friend ontology, a Resource Description Format, and a Web Ontology Language; and
performing a hold on data that a user has access to, wherein the user and the data are identified using the matched user representations for that user.
5. The computer program product according to claim 4 , wherein a non-phonetic algorithm is any one of a q-gram algorithm and a suffix algorithm, and wherein the non-phonetic algorithm is used to group the normalized information sets.
6. The computer program product according to claim 4 , wherein a longest common sub-string is determined using dynamic programming algorithms and is used to determine the similarity score.
7. The system according to claim 1 , wherein determining the similarity score comprises using a longest common sub-string algorithm to generate an output and using any of an Overlap coefficient and a Dice coefficient on the output of the longest common sub-string algorithm.
8. The computer program product according to claim 4 , wherein determining the similarity score comprises using a longest common sub-string algorithm to generate an output and using any of an Overlap coefficient and a Dice coefficient on the output of the longest common sub-string algorithm.