We have several case studies that show you what we have accomplished so far.
The Dutch central government and web texts
Problem definition
Web texts must be comprehensible to the visitors of a website. Therefore, it is importantfor the Dutch government to write texts in simple Dutch language. The editors of Rijksoverheid.nl, the official website of the Dutch central government, use language level B1 as the starting point for their web texts. Our client, a government institution, wanted to know to what extent the texts on Rijksoverheid.nl complied with these guidelines and criteria.
Solution
QA Metrics built a classification tool, based on Artificial Intelligence, that checks to which extent the texts comply with the guidelines set by the Dutch government.
Result
Texts that do not comply with the B1 standard can easily be identified and rewritten quickly. Besides that, new content and text can easily be checked before being published.
Consistency checker for healthcare
Problem definition
Every year, healthcare institutions draw up thousands of treatment plans. Our client, a healthcare institution, desired to know how consistent their treatment plans were in comparison with previously drafted treatments plans with regard to complaints, diagnosis and other relevant data.
Solution
QA Metrics developed a consistency checker based on artificial intelligence, that identifies corresponding and consistent treatment plans within a database.
Result
As a result, the healthcare institution has saved a great amount of time in composing new treatment plans. In addition, the precision and consistency of the treatment plans has been increased.
Anonymization in the legal sector
Problem definition
In the legal sector, court rulings, contracts en personal dossiers have to be anonymized before being archived, which is a costly and time-consuming occupation.
Solution
We developed an AI-based anonymization-tool for the Dutch Department of Justice that identifies data which is subject to the GDPR and automatically anonymizes this data in accordance with the GDPR.
Result
Court rulings and dossiers that have been closed and need to be archived in a database can be automatically anonymized. As a result, the Dutch Department of Justice saved a significant amount of time in anonymizing these documents. Furthermore, the tool has improved the accuracy en reliability of anonymization.