The logic behind 360Science matching engine powers all of our products (mHUB, mSQL and mDesktop).
In order to effectively match contact data, you need an engine built from the ground up to deal specifically with issues not presented in any other form of data matching.
CRM and customer data is unique. The logic behind 360Science matching engine (mAPI) is designed specifically to deal with the nuances of contact data quality issues. Issues not presented in any other form of data matching.
Our team of engineers, data scientists and developers have researched, tested and tested again numerous algorithms from Jaro-Winkler and levenshtein to Spedis, Compged etc to create and perfect the industry’s most accurate and reliable contact data matching engine.
The 360Science Matching Engine – maximizes legitimate matches, minimizes false matches and gives you greater control of the matching process. To quote one of our customers – “360Science’s Matching Engine achieves results that are like human perception”.
To effectively match contact data, you need to understand the context of the full contact record – not just a field. The Matching Engine identifies potential duplicates by taking a 3-dimensional view of the data, never relying on any single item of data being correct or consistent! It does not just compare field to field, but instead it internally identifies, isolates and groups information into their logical parts (such as name and address elements.
To process the typical hearing, reading, keying and “lack of standards” errors found in most databases, we use multiple approaches to ensure that differences arising from all these causes are identified – ultimately finding matches that would otherwise go undetected:
Bower and Bauer
Hernández and Hernandes
Muhammad, and Mohamed
Street & St, Straße & Str.
Auto, Motors and Car
1 = One, First , 1st
Turner Broadcasting Company ~ TBC
LLC = Limited Liability Corporation
Wilson, Wislon & Wilsn
95128 ~ 91528
7350 ~ 07350
Jose Gonzalez
Gonzalez Jose
560 Main St Ste 106
Suite 106 560 Main St
Inc, Incorporation, LLC
Insurance, Assurance
The State Farm Insurance Company of California LLC
The scoring stage determines the actual likelihood of two records being the same.
The mAPI scans through the record, field by field, and works out how similar they are.
#secretsauce
Michael = Mike Michel, Mickey, and Mikhael
Jacqueline = Jacklynn, Jaclyn, Jackie
|Mr Jose R Gonzalez Jr MMD | =
|Mr | Jose | R | Gonzalez | Jr | MMD |
| AtlantaGA30305 |
| Atlanta | GA | 30305 |
ਬਲਰਾਜ = Balarāja
にこらす = Nikkarasu
The mAPI is comprised of three separate distinct processes; (key generation,comparison, and scoring) assembled into a common engine.
You have the ability to create/store persistent keys. (which translates into significantly reduced processing overhead on large data sets.
Your matching logic is independent of key generation, which is extremely configurable and allows multiple levels of match (e.g. individual, business, family and address) to be identified in one pass.
Scoring gives you Control, Visibility and Granularity into your data matching. With the 360Science Matching Engine – you have a highly configurable scoring engine tightly connected to the matching logic – resulting in significantly greater confidence in your matching…
Independent testing by one of our partners reported, 360Science matching engine delivered up to a 226% more accurate match rate on customer CRM data than competing solutions.
No ETL or address standardization needed prior to matching. Yeah – you read that right.
Unlike competing applications or scripted SQL queries – the mAPI doesn’t require data standardization, correction and manipulation prior to matching. That’s because the mAPI isn’t simply comparing field to field. The Matching Engine identifies, isolates and groups contact data into its logical parts. It identifies and compare address and name “elements” as a whole, and allows other data items to be compared in isolation.
The matching Engine also treats address lines as an object. It does not rely on corrected, standardized or normalized addresses to effectively locate matches. And it doesn’t require the same number of address lines in each record to compare the addresses.
This is one of 360Science’s greatest strengths over competing products – especially when requirements are to compare across two or more data sources.
Take a look under the hood and read about the inner workings of mHUB.
Read more about the engine that makes our products #Awesome
Data security and data breaches have become a fixture in the news. You might even recall the story of Dropbox being…
