Automatic handwriting recognition in electronic contract management
Published: on by Khalil Ouardini
Dilitrust offers software to identify and analyze the relevant information contained in your contracts, thus facilitating the digitization of your documents.
Most of the methods we use are based on automatic natural language processing (NLP). In this article, we also show how to use the latest algorithms in this field, in order to optimize the analysis of contracts and improve our NLP tool if necessary.
First, the documents (often scanned beforehand) must be imported into the platform, and an optical character recognition (OCR) algorithm is responsible for extracting the text from the scanned images.
In addition to document processing, it is also necessary to be able to correctly identify non-textual sections, such as signatures or initials. Since these elements are common, we propose a method for detecting signatures on paper documents, presented as a practical guide.