chancegray / id-match-engine

Id match grails app based off of Oyster Match Algorithm Libs

#Date: Sept 2012 #Author: [e-mail] #Desc: id match grails app install/run/test instructions #ref: https://spaces.internet2.edu/display/cifer/ID+Match+Engine

#Dependencies

1. Grails 2.0 ( http://grails.org/ )
2. Java 1.6+
3. Needs MySql 6+ or any other RDBMS

#Setup

1. cd to the folder where you find this README.txt
2. cd grails-app/conf

A.Aegistry Setup

1. Edit DataSource.groovy to point grails app to the database that contains the Person Registry Note: Current version supports only one table.

B.SchamaMap Setup

1. Edit Config.groovy,
2. left-side is for external facing request attribute names, right-side is for internal database column names. ex: idMatch.schemaMap = [ fName : 'firstName', lName : 'lastName', dob : 'dateOfBirth', ssn : 'social' }

C. Match Rules Setup

1. Edit Config.groovy
2. For each attribute that is part of the search filter, add a rule, assign scores for exact and like matches. provide algorithm to use for like matching. ex: idMatch.ruleSet = [ ssn : [exactMatchScore:"50", likeMatchScore : "40", algorithm: "EditDistance", distance:"2"], fName : [ exactMatchScore:"20", likeMatchScore : "15", algorithm: "Soundex"],
   lName : [ exactMatchScore:"30", likeMatchScore : "25", algorithm: "Soundex"],
   ]

D. Cut Off Score Setup

1. Edit Config.groovy
2. Provide cutoffs for exact match and recon match. ex: idMatch.cutOffScoreMap = [ exact : '100', recon : '80' ]

#Search

1. Curl CANONICAL: curl -X POST -d "{"data": {"fName": "venu", "lName": "alla", "ssn": "111222333", "dob" : "123456", "city" : "Berkeley"}}" -H "clientId:tester1" -H "password:123456" -H "content-type: application/json" http://localhost:8080/dolphin/canonical/getMatches FUZZY: curl -X POST -d "{"data": {"fName": "venu", "lName": "alla", "ssn": "111222333", "dob" : "123456", "city" : "Berkeley"}}" -H "clientId:tester1" -H "password:123456" -H "content-type: application/json" http://localhost:8080/dolphin/fuzzy/getMatches NOTE: pipe for formatting the response { | python -m json.tool }

2.Response is similar to : { "exact": [], "input": { "fName": "venu", "lName": "allo", "ssn": 111222333 }, "recon": [ { "fName": "venu", "lName": "alla", "personMatchScore": 95, "ssn": "111222333", "uid": "1111" } ] }