How to Reduce False Positives and Increase the Accuracy of Found Hits in DPS?
When you use a software to screen your trade partners it is quite natural to see positive hits that are listed but not much related to the company or person you are screening. On the other hand it is also possible that true positive results sometimes may not be shown to the user. A critical task that exporters face daily is reducing false positive matches in denied party screening.
In order to reduce false hits and increase the chance of true and correct matches there are a few adjustments that can be done. If the screening software that you use allows you to do search adjustments then they should be used appropriately.
Let’s focus for a moment on ways of increasing the chance of finding correct matches. You can start by tuning your search criteria. In some cases screening software will support phonetic and/or fuzzy logic screening in addition to exact name and address match. Phonetic, or in other words “sound similar’ search will generally use algorithms such as Soundex, Metaphone or Double Metaphone.
Phonetic search feature works by computing code words for each searched word and then compare these code words with the existing code words of the denied entities. Phonetic algorithms are especially useful to find and detect the variations between the names when they are written in different alphabet. For instance, when the name of “viktor” is translated from Cyrillic alphabet into latin alphabet it could be spelled as v-i-k-t-o-r in comparison with the English version of the same name which would be spelled as v-i-c-t-o-r.
Fuzzy search function, on the other hand, would employ certain algorithms where number of different characters in the searched names and the matching names are computed. So, typically accidentally mistyped names can be caught when using the fuzzy search algorithms.
You can also consider searching the entity name and address separately to increase the chance of finding a true match.
Reducing the rate of false positive hits when you are using software to screen trade partners is important. Some of the actions that you can take to reduce the number of false positive hits include the following:
Adjusting Search Sensitivity:
Some software providers offer a slider bar on the search screen where you can actually select a match rate somewhere between 50% and 100% match. Slider bars generally adjust the sensitivity of the used phonetic and fuzzy logic algorithm. So, if the software you are using provides this feature you might try selecting a higher rate than 50%. Increasing the match rate all the way up to 100% would certainly decrease the rate of the false positive hits. However, the flip side of it, this can also cause missing some of potentially correct matches, especially when the searched name has different variations and if you searched and know only one variation of those.
Excluding common words:
Common words generally occur in the name or title of the entity/individual, such as (group, corporation, technology, bank, community, INC, incorporated, hospital, international, LLC, Ltd). If your try adding these words to excluded word tables, they would be automatically removed from all screened entities going forward. Possibility of false results can be greatly reduced by excluding these words. Similar to adjustment of search sensitivity, there is a disadvantage of excluding common word as well. For example, in countries like China, it is quite common to see sanctioned entity names containing such common names. So, exclusion of those words would also mean your search can miss finding some of the sanctioned entities. We suggest caution when selecting excluded words.
Readjustment of what exact lists need to be used:
It is highly recommended that a risk analysis should be done to determine which lists are needed to be used. Type of the business transactions and geographical factors would impose consideration of several other lists. For example, organizations exporting to or doing business in another country should also include that country’s government issued watch lists as well.
Companies that are involved in specific business activities should also use the lists that are issued by the agency regulating those transactions. For instance;
-General Services Administration (GSA) Excluded Parties List would be added to screening lists if you are involved with the Government procurement and contracts;
-FBI and Interpol wanted lists if your business involves the law enforcement or background search;
-Food and Drug Administration (FDA) and Health and Human Services (HHS) banned and prohibited practitioners list if you are in medical, pharmaceutical and other healthcare related sectors.
-World Bank and Inter-American Development Bank lists if you operate in financial sector.
Web based Linqs GtradeProTM SPS Automates Your Denied Party Screening Process. Learn How
Contact Us for More Information