Avid Smart Screen employs a multi-algorithm approach to elevate fuzzy logic matching to deliver high quality matches.
Smart Screen analyses the structure of each name component to perform sophisticated observations using advanced linguistic algorithms.
Hybrid Algorithm Approach
Smart Screen name screening software uses a combination of multiple algorithm methods including a cascade approach to address the maximum number of name variations, utilising:
- Statistical similarity method;
- Edit distance method; &
- Common key methods.
Each of these methods excel at solving one or several of the many challenges to accurate and consistent matching:
- Similar Names
- Out of Order Components
- Missing Components
- Missing spaces & Hyphens
- Split Database fields
- Phonetic Similarity
- Spelling Differences
The first algorithm deployed is to 'de-noise' the data.
When data is ingested the algorithms de-noise the data by replacing noises with alphabetic characters.
Special characters are non alphabetic or numeric characters such as punctuation marks or symbols and are replaced or escaped to ensure they do not influence the results unnecessarily.
With cleaner and more reliable data Avid Smart Screen commences simultaneous actions to identify possible matches in the data.
Multi-algorithm combinations in our hybrid approach miss fewer matches and reduce the number of false positives.
Avid Smart Screen then analyses the data using Similarity Logic and Fuzzy Logic algorithms.
At the same time, Avid Smart Screen calculates the similarity score measuring the words in the search in its entirety and separately, a score for each word contained in the search string and provides a score.
Smart Screen multi-algorithm combination will accommodate:
- identify missing spaces, or hyphens
- transliteration spelling differences
- missing name components
- names split inconsistently or in separate fields.
The inclusion of the additional multi-combination algorithm maximises the precision and usefulness of the results.
Without the right tools screening can be cumbersome, time consuming, increase labour cost and can be fraught with risk in itself.
An effective adoption of technology should include expertise from operations, compliance and technology to refine alert identification.