Using Multiple Algorithms to Improve Organization Name Matching

Posted on July 22nd, 2020

Are the data assets on which your company operate of questionable accuracy and efficiency? We have very little control over how customer, prospect, or other textual data is collected at its source. Therefore there is a good chance you are working within an organization where data quality could be significantly improved, thereby improving every process, activity, or analysis that makes use of this data.

There are many examples of data quality challenges for an organization to overcome. One is the difficulty of dealing with inconsistent company/organizational name data, such as the following:

IBM, International Business Machines, ibm, I.B.M. Forever 21, Forever Twenty-one, forever21 G.E., General Electric, Gen Electric Johnson & More, Johnston and Moore Attys at Law Apple, The Apple Store, Apple Corp USA San Jose Memorial Hospital, Med Center of San Jose (Mem Hospital)

Inconsistent company/organization names like these in your databases can cause significant problems, such as:

- duplicate company records
- data and contacts spread across multiple records
- inaccurate data analysis
- difficult to match across data sources
- faulty-decision making
- increased data management costs
- inefficient business processes
- missed opportunities

Interzoid has created a similarity key generation capability that greatly simplifies and can even solve these challenging data issues.

Our API generates a similarity key from a company name that is algorithmically designed to eliminate the inconsistency in the data content. Here are some examples of generated similarity keys:

Similarity key generation example for a company name matching API

We have recently added the ability to utilize different matching algorithms when generating similarity keys. A "wide" key generation algorithm casts a wider net of possible matches for a given organizational name. The tradeoff here is an increase in the possibility of false positive matches where organizational names are similar. The "narrow" algorithm tightens the match requirements for two entities to generate the same similarity key. However, because of these tighter requirements, some matches could be missed. Of course "medium" is somewhere in the middle. So depending on your organization's requirements and business case, there are tradeoffs to consider.

Once similarity keys are generated for each organization in the target data source, data can be matched or sorted using these similarity keys rather than the data itself, resulting in dramatically higher match rates and identification of duplicate company names.

Using Interzoid's Company Name Match API, you can:

✔ Eliminate redundant and duplicate data from customer and important databases
✔ Leverage data content-specific matching algorithms for higher matching success rates
✔ Improve the accuracy of data analysis activities with better, more accurate foundational data
✔ Be more cost-effective with customer communications and marketing campaigns
✔ Utilize fuzzy matching to match company data across datasets
✔ Enable fuzzy searching for better, more comprehensive search results
✔ Achieve greater ROI with business intelligence investments
✔ Leverage growing AI-based knowledge bases of matching and ambiguous data identification
✔ Reduce costs associated with redundant, duplicate data
✔ API-based solution enables full customization of data matching strategies
✔ Similarity keys easily are leveraged within a broad range of applications
✔ Easy to get up-and-running with immediate results
✔ Pay-as-you-go pricing makes it easy to get started

Sample Matches Found Using Similarity Keys:

IBM International Business Machines i.b.m. corp U-Haul Uhaul Trailers Costco Stores cost-co McKeson Corp Mackesson The Mckesson company Florida St Florida State University FSU Well's Fargo Bank Wells Fargo Nationwide Insurance Nation Wide Ins. PFIZER INC Fizer Forever 21 Inc. FOREVER TWENTY-ONE Biogen Pharma The Biogen Corp Apple apple computer apple comp APPLE MOBILE Apple International apple inc apple inc. Apple Stores THE APPLE STORE Apple Corp Apple USA 7-11 7-eleven Seven Eleven Stores Seven 11 Inc. Hilten Hotels HiltonHotels Hillton Resorts Inc. Arctic Flooring Inc Artic Floors USX Corp United Steel Plain Dealer Cleveland Plain Dealer cleve Plain Dealer Cleaveland Plane Dealer Glaxo Smith Klein GSK Pharmaceuticals San Deigo Chemical SD Chem San Diego Chemical Corp 10k investments Ten K Investment Jackson Wines Jaxon Winery Inc. 5 Penn Plaza Inc. Five Pennsylvania Plaza Cranston Inc. The Cransten Corp Cranstin Company Hilton Hotels Hillton Resorts Inc. Perez Telecommunications Inc peres telcom Mary's Cookies Marys Cookie Shop High Top Brewing Hightop Breweries The Johnson Rehab Center Johnsen Rehabilitation JP Morgan JP Morgan Chase JPMorgan Chase Midland Loan Services Mid-land Savings & Loan Midland Banking Svcs Equity Real Estate Investments $Equity Realtors NBA National Basketball Association National Basketball Assoc. (NBA) Saint Francis Mem (main) St. Francis Memorial Hospital

See our Snowflake Native Application. Achieve Data Quality built-in to SQL statements.

Identify inconsistent and duplicate data quickly and easily in data tables and files.
More...

Connect Directly to Cloud SQL Databases and Perform Data Quality Analysis

Achieve better, more consistent, more usable data.
More...

Try our Pay-as-you-Go Option

Start increasing the usability and value of your data - start small and grow with success.
More...

Launch Our Entire Data Quality Matching System on an AWS EC2 Instance

Deploy to the instance type of your choice in any AWS data center globally. Start analyzing data and identifying matches across many databases and file types in minutes.
More...