Depending on your organization’s use case or needs, you may require either very strict or very liberal matching parameters. Persona’s customization process is granular and detailed in order to allow our customers more control over what matches they receive, and how often they receive them (false positives).
Deciding what match requirements work best for your business can be intimidating - it’s like a whole new language! We’ve put together this help center article to help you master the terminology so that you can feel empowered to determine the best match requirements for your business. You can also check out our blog post on match requirements to learn even more.
Learning more about your match settings, and how to control them could help decrease the number of manual reviews that your team has to complete on a daily basis, or could help you cast a wider net to ensure you’re covering all your bases.
As outlined in our help center article, Reports Template Editor Overview, match requirements are now configurable through the Persona dashboard, provided you have the right permissions to do so.
Under the Name matching section in the Reports Template Editor on a per Report basis, you will now be able to see the existing match requirement settings. Please see terminology listed below to understand exactly how you may want to configure your settings.
The following table shows examples of how the options are used:
Exact name: exact name takes the primary name and determines whether there is a match. Primary name refers to the name that is considered a hit in the article or website. There are no allowances for typos in an exact match setting. In the Reports Template Editor, this is always selected by default so you can receive exact matches.
- Example: Search term = John Smith, result = John Smith only
AKAs: AKAs setting includes exact name match, in addition to an alias. An alias is another name that the person is known by. This is still an exact match. There are no allowances for typos.
Partial names: partial names setting includes partial name matching in addition to AKAs and exact name match.
- Subset name matches: a match on a subset name matches on all of the subset of the words.
- Example: Search term = John Smith, Profile matched = John Bob Smith
- Initial matches: a match on an initial matches on entities that only have an initial for one or more of the words.
- Example: Search term = John Bob Smith, Profile matched = J.B. Smith
Equivalent names: equivalent names setting includes partial names, AKAs, and exact name matches. Equivalent names also includes alternates for the name passed in. This can included hypocorisms and homophones.
- Hypocorism: nicknames
- Example: Search term = Bob Smith, Profile matched = Robert Smith
- Homophones: names that sound the same as the matched name
- Example: Search term = Jon Smith, Profile matched = John Smith
Joined matching: attempts to match by comparing only the letters in the order provided (ignoring punctuation and whitespace).
- Example: Search term = John Mc.Smith, Profile matched = John McSmith
First/Last matching: attempts to match by comparing only the first and last terms.
- Example: Search term = John Jacob Jingleheimer Smith, Profile matched = John Smith
Fuzziness: fuzziness has many different components and is a custom setting. For the majority of Report types, if typos are allowed, the types of typos allowed are phonetic substitutions - meaning a typo on a vowel.
- Watchlist when "Partial Names" is NOT selected
- Low - 1 typo allowed on entire name at a minimum of 18 characters
- Medium - 1 typo allowed on entire name at a minimum of 10 characters
- High - 1 typo allowed on entire name at a minimum of 3 characters
- Watchlist when "Partial Names" is selected
- Low - 1 typo allowed per word at a minimum of 18 characters
- Medium - 1 typo allowed per word at a minimum of 5 characters
- High - 1 typo allowed per word at a minimum of 3 characters
For the Watchlist report specifically, fuzziness and typos can also include additions, subtractions, and substitutions. Below are examples based on the original name: Thomas Henry Jefferson
Addition: means that an extra letter is mistakenly added to a name.
Match addition example #1: Thoomas Henry Jefferson Match addition example #2: Thomas Henryy Jefferson
Subtraction: means that a letter that should be in the name, is mistakenly removed from the name.
Match subtraction #1: Tomas Henry Jefferson Match subtraction #2: Thomas Henry Jeferson
Substitution: means that a letter in a name is replaced/substituted mistakenly for another letter.
Match substitution #1: Thomis Henry Jefferson Match substitution #2: Thomas Henri Jefferson