Predictive Justice (PJ) is an ambitious long-term project built on an innovative approach and philosophy. We start from the assumption that only combining different expertise and tools in coherent and integrated pipelines brings to effective results. For this reason, we organize our work modularly, in connected but  scientifically and operationally autonomous projects -- each tackled by an interdisciplinary team. PJ has multiple and transversal aims: from the attempt to “export” knowledge, techniques, and solutions across disciplines (e.g., from omics to legal data mining), to the coupling of protocols and software to automate the pseudonymisation of texts, to the creation of innovative tools for querying legal materials through their automatic annotation, to the construction of predictive tools based on Data Science and Artificial Intelligence, to the attempt to offer comprehensible explanations on the functioning of the tools used and adapt them to end-users’ needs and abilities. All these aims are articulated in full coherence with the corresponding regulatory and ethical frameworks, in the belief that regulatory and ethical profiles are key to our research and go beyond mere adherence to norms.  PJ activities cut across the LiberLab (Giovanni Comandé, Denise Amram) and EMbeDS (Daniele Licari, Francesca Chiaromonte).

THE MAIN CHALLENGES OF PREDICTIVE JUSTICE

• Create a robust, ethically and legally compliant infrastructure to collect and archive case law decisions using innovative tools.
• Simplify access to the legal records, automate case law research activities, and thereby reduce the costs of the justice system.
• Harmonize and anonymize data streams to facilitate reuse and sharing.
• Increase understanding of legal processes and evaluation in legal disputes by identifying possible guidelines.
• Create a network of lawyers, judges, doctors, data scientists, social scientists, etc. to increase the involvement of institutions and stakeholders to better understand the justice system and our society.
• Explain the reasoning behind each decision for different stakeholders.
• Provide support in drafting scales in some civil litigation.
• Draw the boundaries between interpretive consistency and unpredictability of justice.