PNA conducted a comprehensive study on the responsible support of legal analysis classification. This research focused on the problem of potential errors and wasted time in performing manual tasks within Law analysis by analyzing the structure of laws. This was done using the legal reference model (JRM 2.0).
Currently, we are in the process of implementing of the JRM Annotation assistent in the iKnow Cognitatie module of PNA, whereat users can select text and receive suggestions on how it can be annotated. In addition, a probability score is added to indicate how certain the model is about the appropriateness of the annotation.
Although all of this is powered by artificial intelligence, the user remains in control at all times. Each suggestion must be approved. Users can, of course, also make changes to the suggestion based on their knowledge of the specific domain. Finally, they can reject the suggestion if it is not deemed appropriate.
The system remembers these choices and collects them as training data for future cases, leading to further improvement of the model and thus an increasingly efficient use of the user’s time.
The efforts so far have resulted in a functioning assistant that can intelligently support knowledge modeling and legal analysis professionals. The JRM annotation assistant helps them by improving data collection, enabling knowledge reuse and promoting higher productivity. Continuing this work will only produce better results in the future, further increasing these benefits in daily practice.