Funding by the European Union (ERDF)
Applications, informal letters, invoices, certificates and receipts… The sheer number of documents in a company ties up important resources. Letters, emails and applications must be read and forwarded. Important information about senders and concerns is manually transferred to databases before a clerk can deal with the actual content.
To avoid such overhead, PLANET AI is developing an assistant system for document processing based on deep artificial neural networks in the project Automatic Information Extraction from Documents (AID). A complex task such as information extraction must combine textual and visual features to achieve an optimal result. PLANET’s award-winning handwriting recognition and layout analysis technologies form the basis for these features. State-of-the-art technologies from the field of Natural Language Processing are being further developed so that they combine features and extract information with high precision.
The customer will be able to train systems themselves and adapt them to their needs without specifying complicated rules or templates. The costly data production is also to be largely avoided, and the assistant should be able to learn new classes and features from just a few example data points.
The project duration is from 2019 to 2021.
