What is Intelligent Word Recognition?
Intelligent Word Recognition uses artificial intelligence to recognize whole words in a document instead of individual characters (which is how Optical Character Recognition / OCR works). For example, when an OCR system is extracting the word “dog” from a document, it will recognize “d”, “o”, and “g”. IWR will match the letters to a dictionary and extract the whole word, “dog” based on pattern recognition and matching algorithms.
Captricity delivers more than both approaches.
Intelligent Word Recognition does not recognize names and numbers not housed in a catalog or dictionary. While Captricity uses OCR algorithms as a first pass on every data capture job, we have tweaked our algorithms with machine learning and pattern recognition to deliver higher results than standard OCR systems. But what really sets Captricity apart from IWR and OCR is crowdsourcing. To ensure accuracy, every data field is verified by between one and five people, enabling our stellar 99% accuracy rate.
Captricity and crowdsourcing
Captricity uses crowdsourcing to verify the data contained in every field we capture. When customers set up their initial job, they upload a blank version of the document they need extract information from. They then use our system to delineate which fields need to be digitized. Each form is “shredded” into distinct fields and each field is verified by one or more people. Since no one person can see more than one field from each document, privacy is maintained.