How does Skandy work
If the image is uploaded
First, Skandy uses the Optical Characters Recognition (OCR) technology to identify all the text in the image. At the time of uploading a file, it is required to select the language of the text in the image properly in order to recognize the text accurately.
If a text or text file is uploaded
Using algorithms, the text is split into logical parts. The received parts pass through Microsoft and Google services in order to find matches on the Internet and public databases. Text fragments are checked not only ‘as is’ by search engines, the number of consecutive words that match, the distance between words, the length of words, the probability of changing, replaced and deleted words and even more — are taken into account.
The final resulting percentage calculated by the formula:
[Number of excerpts matched] / [Total number of excerpts],
where an excerpt can be not only a whole sentence, but also a phrase of sufficient length with the necessary meaning.