FuzzyDupes v4.4 full

Did you know that your contact database typically contains 3-10% duplicates ?
These duplicate records result in unnecessary costs when sending out printed catalogs, are aggravating to your customers, create problems in the controlling, etc. With classical methods you have no possibility to locate these duplicates in your database.
With it's highly optimized pattern matching algorithms, FuzzyDupes finds similar records with high accuracy and precision.

FuzzyDupes searches for similar records – especially in address databases.
Although fuzzy Pattern-Matching Algorithms use up a lot of computing power, this program can manage, for example, 30,000 records in about 20 seconds*.

The duplicates are then displayed in a clearly arranged table. You’ll be impressed by the results of the duplicate search.

FuzzyDupes processes the data in memory and implements an intelligent procedure of Pattern-Matching, excellent in quality, selectivity and performance. With our Trigramm/Hashindex, the component finds small and reliable clusters in the first step. Then, assembler optimized pattern matching algorithms test the records for similarity with mathematical exactness.

*) with P4, 2.4 GHz, 256 MB Ram

Up till now, duplicate comparison was a task for expensive applications and specialists. FuzzyDupes allows small and mid-sized companies to remove duplicates from their databases, thereby reducing costs (e.g. shipping costs).




Supported Data Sources / DBMS:

MS-Access
MS SQL-Server
MS-Excel
Text/CSV Files
Other Datasources with ODBC-Driver, e.g. Oracle, MySQL, etc.
MS-Outlook
Windows Addressbook

Features
Fast fuzzy duplicate search in many data sources
Fuzzy merge of two lists
Fuzzy match with external list (Robinson list)



New Features in version 4.0
Version 4 contains our new computing kernel which is highly improved in speed and quality.

The duplicate search runs completely in the RAM and is up to 100 times faster than in earlier verions
You can search in any data sources: All ODBC databases, Text/CSV Files, MS-Outlook, Windows Addressbook
A new method for cluster creation gives mathematical exact clusters in a shortest time
The computation kernel is highly optimized implemented in C++

System requirements:

Win95 / Win98 / ME / NT 4.0 / Win2000 / XP

See more on developers site.