Correct your data automatically
Thanks to its on-board artificial intelligence engine, Tale of Data boasts a hundred or so functionalities designed to improve data quality.
To correct your data effortlessly, without writing code, you can use :
-
ready-to-use functions for automatic error correction,
-
fuzzy logic, to create bridges between similar data and make your information more reliable,
-
repositories to complete your information,
-
transformations to normalize or deduplicate your data,
-
business rules in natural language to apply your standards.
Use ready-to-use functions
Tale of Data features a large number of ready-to-use drag-and-drop functions.
Among these functions, the ones most used by our customers :
-
deduplication with confidence index,
-
automatic detection of standardized fields: telephone number, SIRET, ISO code (country, etc.), NIR code, ....,
-
suggested corrections
-
fuzzy logic to reconcile "almost" identical textual data.
Automatically correct "close" data
Using fuzzy logic, or Fuzzy Maching, Tale of Data detects similar data.
The power of fuzzy logic functionalities makes it possible to match identical or similar elements written in different spellings.
Select the strategy that corresponds to the data to be corrected.
-
phonetics: to reconcile different spellings of the same sound (o, au, eau, etc.),
-
consonants first: this technique is based on the principle that consonants are more important. These letters are therefore weighted more heavily and examined first for duplicates,
-
vowels first: same idea, but on vowels,
-
N-gram: this technique divides strings into smaller sub-strings, called "n-grams", which are then used to detect similarities.
Enhance your panel of preset corrections
Use repositories to correct your data
Although the main reason for using repositories is data enrichment, this Tale of Data feature is also used to correct your data. In fact, you can compare your data with reference data and correct it in just a few clicks.
Correct your data with your own business rules
Adding your own rules, in natural language* (drag & drop), ensures that your data complies with corporate rules and standards.
Example of a business rule: IF the product is in category "C" AND the order contains at least 5 items THEN apply a discount of 4%.
