Augmented Reading and Similar Case Matching: from Legal Domain Experts’ Modus Operandi to a Computational Pipeline

The increasing backlog in processing legal cases in the Courts pushes toward the implementation of solutions that can reduce the workload and thus the time to complete cases. Searching for similar judgments is a relevant task that enables consultation of related legal proceedings. This paper proposes to automate the search process by developing a pipeline inspired by the modus operandi of legal domain experts. The pipeline includes an augmented reading of judgments for the purpose of semantically analyzing texts taking into account the context given by their classification, decisions and citations. The results are interpretable by legal domain experts, while the similarity case-matching output is underpinned by the information extracted from the relevant documents. The paper addresses a case study based on Italian national laws.