The steadily increasing workload per employee is visible in almost all areas of work. Mass proceedings are becoming an ever-increasing problem for judges (see statement by the German Association of Judges ). Examples of mass proceedings include claims for damages for diesel vehicles or the assertion of air passenger rights. Online platforms and specialized law firms are responsible for an ever-increasing number of lawsuits in an ever-shorter period of time. In addition, the same pleadings are often used without considering the individual case. This results in extensive pleadings with many attachments and masses of pages where only a small part of the attachment is relevant to the case. Judges have to work through all of this within set deadlines. However, as the cases from the mass proceedings are usually similar, the level of monotonous work increases in addition to the workload. A combination that challenges and burdens the judge.

Artificial intelligence (AI) can help here by facilitating process management. Since mass processes always involve the analysis and further processing of texts, AI techniques from Natural Language Processing, or NLP for short, can be used here. The field of NLP deals with the understanding and processing of language in texts. AI can be used for different purposes. On the one hand, the AI can read texts and write a summary at the end or filter out information from a text and assign it to categories (Name Entity Recognition – NER). With its STAN(SemanticText Analysis) project, Westernacher Solutions has specialized in extracting information from legal documents. This allows PDF files or scanned documents to be divided into categories (lawsuit, judgment, …) and procedural information to be extracted from documents. For example, parties involved, their names and addresses, the amount in dispute, subject area and file number can be recorded. All of this information can be used to direct documents received by the court to the relevant files, databases or judges. Judges then receive all the necessary information on the new documents in a brief overview.

This AI support is an enormous relief, especially when it comes to preparing private lawsuits or lawsuits via online platforms. This is because the XJustiz data format used for electronic legal transactions, in which all information is normally available in digital and structured form, is not used here. Document data, process data and process information would then have to be created manually in these cases. With the use of AI, this monotonous and time-consuming step is automated. The case recording process can thus be made much faster and more efficient.

The extraction of process information has yet another advantage. Judgments, e.g. civil court decisions, must be published anonymously. Within the STAN project, all information from legal documents, including judgments, is recognized and marked. This can be used as preparation for a further step, namely the anonymization of the judgments. As a first step, all information can be redacted and thus made publicly available. In a second step, the necessary information can also be replaced by fictitious values. The text content remains the same, but the judgment has been completely anonymized. The advantage of this step is that the judgments can be used for further processing without losing information or text contexts. This is particularly important in the field of NLP for the successful analysis of texts.

Anonymization therefore not only serves the purpose of the obligation to publish judgments, but also as preparation for further AI technologies and solutions.

Westernacher Solutions thus not only offers an opportunity to speed up the work of judges through mass proceedings, but also to pave the way for more AI solutions.

In the spirit of “Go Digital. For Sure.