Improving semi-automated export check with a content based approach
Date:
This talk was a presentation of the semi-automated output checking system in its revised 2025 version. It aimed to answer questions about :
- What is the best way to represent files in order to analyse them ?
- How can be build efficient features to assess and predict the level of risk carried by a file ?
- How do we honestly evaluate the performance of such a model ?
- General recommandations and observations about an export system, to make it easier to automate