Intelligent document classification and interpretation
| Reference number | |
| Coordinator | Qoorp AB |
| Funding from Vinnova | SEK 500 000 |
| Project duration | May 2016 - January 2017 |
| Status | Completed |
Important results from the project
The purpose of the project was to extract actionable intelligence from business documents. The particular problem we chose to solve was to recreate the shareholder register from public documents submitted to the Swedish Company Registration office for other purposes. By combing that data with data from public APIs, we were able to build a prototype that could, with minimal human input, recreate the shareholder register.
Expected long term effects
Parts of the results have been implemented in our product. This concerns certain algorithms that we developed for the purpose of transforming data. We have also verified the need for the product with certain actors on the open market.
Approach and implementation
We started with an analysis of the types of data that is submitted when registering share issues and other corporate actions that need to be registered with the Company Registration Office. We then analyzed what data was available through APIs and from other sources, in order to determine what we useful information we could extract.