What is the difference between the values of "matchedByDocuments" and "matched"?

Question

The verify matches API we are using shows that the records are MatchedByDocuments true but actual match is false between that.

What could be the possible reason for the same? The API we are using is 

 

POST https://{{env}}.reltio.com/reltio/api/{{mdmTenant}} /entities/_verifyMatches

 

Answer

 The "matchedByDocuments" field indicates whether the entities are matches according to the comparison scheme (operands + comparators). The matched field indicates whether the entities are matches according to the comparison AND tokenization scheme (there is an intersection of match tokens by the match rule).
 
Look at the following in the response.  You will notice that the intersection field is empty.
 

POST https://{{env}}.reltio.com/reltio/api/{{mdmTenant}}/entities/_verifyMatches
?rules=configuration/entityTypes/HCP/matchGroups/HCPRule3c&show=full

 
"matchTokens": {
             "first": {
                 "foundInMatchTables": true,
                 "tokens": [
                     "J7uB1QHAbCM="
                 ],
                 "tokensGenerated": 1
             },
             "second": {
                 "foundInMatchTables": true,
                 "tokens": [
                     "+Emg208Nsfk=",
                     "NNDOJU/6RTo=",
                     "GVORnsqHRAo=",
                     "8095lcAwwRE="
                 ],
                 "tokensGenerated": 4
             },
             "intersection": {}
         },

 If there are no common match tokens the entities are not identified as match candidates and are not compared.
 
Ways to resolve:

  • Nothing to fix. It should be left as-is:
    • the entities are different
  • Check the tokenizer assigned to the attribute. FirstName has DynamicDamerauLevenshteinDistance comparator but ExactMatchToken match token class. The match token class does not correspond to the comparator class. The behavior, when the comparator class is looser than the match token class, is usually wrong. The match token class has to identify the same or looser set of match candidates than the set of matches identified by the comparator class. In this case, the solution might be another match token class that allows some digits to differ (see CustomMatchToken).
  • Add the attribute to ignoreInToken.
Was this article helpful?
0 out of 0 found this helpful

Comments

0 comments

Please sign in to leave a comment.