Data merging 4.0: Use of fuzzy algorithms

Large amounts of data from many different and unlinked data sources are a real challenge in data processing. Often there are no unique keys for linking in different data sources. Furthermore data records that are actually the same easily have different spellings & misspellings, missing information or double entries (duplicates).

Therefore, e.g. address matching of different data sources is often only possible with a lot of time (often also manual checks). Due to this reason matching tasks are only carried out when absolutely necessary.
But: it is precisely well-linked data that brings the real added value for analysis and practical use by the staff!

Since we are confronted with the issue of data merging / matching on a daily basis in our projects, we have taken our existing matching procedures to the next level. We now use fuzzy algorithms that can optimally support us in finding data matches. Of course, detailed cross-checking is still done manually for quality assurance.

In this way, we regularly achieve very high assignment rates, which then provide a high added value for further processing and our analysis questions.

So if you also despair of your unlinked data, then contact us directly!