MerlinMerge® SpeedPro Uses “Fuzzy” Logic to Improve Searching and Matching

MerlinMerge® SpeedPro provides extremely accurate and powerful data deduplication and merge/purge operations. At its core is IST’s powerful searching and matching technology, which allows it to outperform the competition. How? By using sophisticated techniques, such as “fuzzy” matching, heuristic algorithms, phonetic analysis, and much more.

As opposed to exact matching, “fuzzy” matching attempts to improve searching and matching results by being less strict but without sacrificing relevance. The "fuzzy" matching algorithms are designed to find strings related to the terms used in the search. For example, related words are likely to have the same core and differ at the beginning and/or end - Rebecca and Becky or John and Jonathan. Fuzzy matching allows these seemingly different names to match, a desired result that most of us prefer.

The "fuzzy" matching technique is often neglected in searching applications because
people do not realize that exact matching will simply not do the job. For example, when you search for George, a misspelling such as Goerge will never be found using exact matching.

Wildcarding has the opposite result: returning too many matches, many of which are not relevant. For example, searching for Mar* will return Margaret, Martha, Mark, Marlon, Marnie, and Mary when all we really wanted was Marilyn.

Rather than resorting to exact matching or wildcarding, MerlinMerge® SpeedPro uses sophisticated approximation techniques to convert many different near-miss situations (such as those involving faulty prefixes or suffixes, character misplacement, nonstandard word stems, etc.) into more adequate results. While sophisticated "fuzzy" matching algorithms often sacrifice performance, MerlinMerge® SpeedPro applies the "fuzzy" logic only to a subset of previously filtered relevant records.

It is important to understand that MerlinMerge® SpeedPro does not rely on "fuzzy" matching alone but interweaves it with advanced heuristic algorithms, phonetic analysis and scoring techniques to achieve unsurpassed accuracy and to reduce the number of false positives (records that were matched but were not really a match).

The sophisticated techniques that are at the heart of MerlinMerge® SpeedPro are hidden from the end user by an easy-to-use, point-and-click graphical user interface. With only a few mouse clicks, users can send their merge/purge or deduping jobs to the matching engine and receive processed results in a specified format.