نوع مقاله : مقاله پژوهشی
نویسندگان
1 استادیار گروه پژوهشی فقه و حقوق اسلامی، دانشگاه فردوسی مشهد، مشهد، ایران
2 استاد گروه حقوق خصوصی، دانشکده حقوق و علوم سیاسی، دانشگاه فردوسی مشهد، مشهد، ایران
3 استاد گروه مهندسی کامپیوتر، دانشکده مهندسی، دانشگاه فردوسی مشهد، مشهد، ایران
چکیده
کلیدواژه‌ها
عنوان مقاله [English]
نویسندگان [English]
The computationalization of legal reasoning is considered one of the most challenging topics at the intersection of law and artificial intelligence.The present study aims to elucidate the substantive distinctions between the paradigms of legal reasoning and algorithmic reasoning, and to propose strategies for their integration.The core hypothesis of this research is that, despite serious obstacles to the complete computationalization of legal reasoning, it is possible to design intelligent decision-support platforms; platforms that preserve fundamental legal principles alongside utilizing technical capabilities.In this regard, employing a descriptive-analytical approach and a comparative method, the study investigates the nature and mechanism of algorithmic reasoning and compares it with legal reasoning.Findings indicate that the normative, interpretive, and contextual nature of legal reasoning, as opposed to the computational, often deterministic, and largely abstract character of algorithmic reasoning, creates fundamental barriers to the convergence of these two reasoning paradigms.These substantive distinctions give rise to challenges that are particularly intensified in the Iranian legal system due to the complex structure of Islamic jurisprudence (Fiqh), the multiplicity and occasional conflict of sources, the specific nature of legal interpretation, and the difficulty of aligning technological dynamism with jurisprudential foundations. Accordingly, the results suggest that addressing these challenges—moving beyond the rule-based reasoning and mathematical-probabilistic approaches that have been most widely used thus far—requires adopting more advanced approaches in intelligent systems.These include the deployment of tree-structured hierarchical agents to ensure the transparency of the reasoning path, as well as the use of hybrid models based on linking general rules to specific cases.
کلیدواژه‌ها [English]
)