In fact, it can be a villain when you have to deal with multiple lists, which is where the unique identifier problem usually pops up. The unique identifier, of course, is just a label, a string of numbers or of letters (which could be a proper name) that designates or denotes one and only one entity or other entry in a particular list.
"Particular list" is the key. In HR, there might be a list of all candidates who have applied, attended a job fair, or been entered from other promising sources and each individual has a unique identifier. The same is true of another list that perhaps you purchase. Or a list of this year's local tech-school graduates. The problem is that each list, if accurate, will use unique identifiers—but not the same identifiers as your list.
That, of course, is the "unique identifier" problem. You want to build your list by adding other useful lists. Then, you want to search the new list with queries to which the entire list responds. Or, you want to use the list to send emails, but not two to the same individual. Everyone has experienced the latter problem: How to get rid of duplicates when the same person on two different lists may have two different unique identifiers?
All this is pretty well known in HR departments, where lists are a way of life and good lists are essential:
- You want to "clean" lists of duplicates and of items that have become out of date (stale).
- You want to merge lists with different sets of unique identifiers and get a new list with where all identifiers are consistent.
- You do not want to do this "by hand," even for relatively short lists, and you can't possibly do it by hand for longer lists.
What you need is a software program that merges lists with inconsistent unique identifiers, in some cases no unique identifiers, and produce a list free of duplicates and stale entries, with consistent identifiers, and searchable by other categories than identifiers: year of entry into the list, sex, relationship with our firm, and so on.
That can be a real problem, as we know. And the solution to it is software system of a kind that at Swooptalent we nicknamed "the secret sauce." There is not a single ingredient in the sauce, brewed over a rather long and expensive development period, but one ingredient worth mentioning is algorithms using "fuzzy logic."
A technical explanation isn't needed, here, but an easy way to understand fuzzy logic is that a unique identifier is exact—like the combination to a safe. Either it is or it isn't the combination. It opens the safe or it doesn't. Fuzzy logic, by contrast, involves rules for searchers and identifications that can locate what is "similar," what is "like," what is related in a particular aspect—not just what is identical. Not just what is the sole and single combination to the lock.
Super-fast reiterated searches using fuzzy logic--and other software methods and systems—keep querying different lists, asking progressively changing questions, imposing tightening parameters, to come up with the list we want. It goes without saying that the result is not 100 percent perfect; only combinations to locks are.
But the Swooptalent system, incorporating sophisticated algorithms for dealing with initially incompatible lists, can enable you to combine lists, incorporate new lists, update legacy lists, and discard not needed lists to produce the result you need.
Often that is done at the flip of a switch. It is artificial intelligence at its best, seemingly endless sorting, checking, and recategorizing to save literally days of productive time. And achieve results that save additional time because new lists make possible targeted searches, directed emails, and reports requested by other departments.
We are proud of the Swooptalent system. We are pleased to answer your questions about it or provide a demonstration. Actually, we view it as a kind of super hero.
No Comments Yet
Let us know what you think