ÃÊ·Ï ¿­±â/´Ý±â ¹öư

º» ³í¹®Àº ÄÄÇ»Å×À̼ųΠ»çȸ°úÇÐÀÇ ¹æ¹ý·ÐÀ» ¹ýÇп¡ À̽ÄÇÏ´Â °üÁ¡¿¡¼­, ÄÄÇ»Å×À̼ųιýÇÐ(Computational Law)ÀÇ °³³ä¡¤¿ª»ç¡¤ÇØ¿Ü µ¿ÇâÀ» Á¤¸®Çϰí Çѱ¹ ¹ýÇÐÀÇ Á¦µµ¡¤±â¼úÀû¿©°ÇÀ» Áø´ÜÇÏ¿© Çѱ¹Çü ½ÇÇà ÇÁ·¹ÀÓÀ» Á¦¾ÈÇÑ´Ù. ¿¬±¸´Â µ¥ÀÌÅÍ ¼öÁýÁ¤Á¦Å½»ö¸ðÇüÈ­Æò°¡·Î À̾îÁö´Â Ç¥ÁØ ÆÄÀÌÇÁ¶óÀÎÀ» ºÐ¼®Æ²·Î »ï¾Æ, ÀÔ¹ý¡¤»ç¹ý ¿µ¿ªÀÇ ¹ý·É¡¤ÆÇ·Ê¡¤ÇàÁ¤±ÔÄ¢À» ±â°è°¡µ¶ ÇüÅ·Π±¸Á¶È­Çϰí Åë°è¡¤±â°èÇнÀ¡¤³×Æ®¿öÅ© ºÐ¼®À» °áÇÕÇÑ ½ÇÁõ °¡´É¼ºÀ» °ËÅäÇÑ´Ù. ÇØ¿Ü¿¡¼­´Â Jurimetrics¿Í Stanford CodeX, MIT Computational Law LabÀ» Áß½ÉÀ¸·Î ±Ô¹üÀÇ Çü½ÄÈ­, °è¾à ÀÚµ¿È­, ÆÇ·Ê ¿¹Ãø µî¿¡¼­ ÃàÀûµÈ ¼º°ú°¡ È®ÀεǸç, ±¹³»¿¡¼­´Â¹ý·É Open API¿Í ´ë±Ô¸ð ¼Ò¼Û ȯ°æ µî ÀáÀç·ÂÀÌ ÀÖÀ½¿¡µµ ÆÇ°á¹® °ø°³ ¹üÀ§¡¤Çü½ÄÀÇ Á¦¾à, Ç¥ÁØ¡¤±â°è°¡µ¶¼º ¹ÌÈí, ±ÔÁ¦¡¤À±¸® ºÒÈ®½Ç¼ºÀÌ º´¸ñÀ¸·Î ÀÛµ¿Çϰí ÀÖÀ½ÀÌ µå·¯³­´Ù. ÀÌ¿¡ ´ëÇØ º» ³í¹®Àº (i) ÆÇ°á¹® ÅØ½ºÆ®È­¿Í »ç°Ç ¸ÞŸµ¥ÀÌÅÍ ½Ã¹ü Ç¥ÁØ, (ii) ¼öÄ¡¡¤ÀÚ°Ý Á᫐ ºÐ¾ß¸¦´ë»óÀ¸·Î ÇÑ ¡°Rules as Code¡± ÆÄÀÏ·µ°ú °ñµç ÄÉÀ̽º ±â¹Ý °ËÁõ, (iii) LLM°ú ±ÔÄ¢¿£Áø¡¤Áö½Ä±×·¡ÇÁÀÇ ´º·Î½Éº¼¸¯ °áÇÕÀ» ÅëÇÑ ¼³¸í°¡´É¡¤±Ù°ÅÁ¦½Ã °­È­, (iv) ºñ½Äº°¡¤¸ñÀûÁ¦ÇÑ¡¤Á¢±ÙÅëÁ¦¡¤¶óÀ̼±½º¸¦ Æ÷ÇÔÇÑ Àç»ç¿ë °Å¹ö³Í½º¿Í °¡À̵å¶óÀλ÷µå¹Ú½º(v) ¹ýÇÐÄÄÇ»ÅͰøÇÐÆò°¡À§¿øÈ¸ÀÇ »ïÁß ±¸Á¶, Åë°è/»ê¾÷°øÇÐÀÇ °øµ¿ ±³À°¡¤Å¬¸®´Ð¡¤ÄÁ¼Ò½Ã¾öÀ» ÇÙ½É °úÁ¦·Î Á¦½ÃÇÑ´Ù. ¾Æ¿ï·¯ 0 12°³¿ù(Âø¼ö), 1-3³â(È®Àå), 3-5³â(Á¤Âø)ÀÇ ´Ü°èº° ·Îµå¸ÊÀ» ÅëÇØ Áö½Ä±×·¡ÇÁ ÅëÇÕ, ÀÚµ¿ Áؼö Á¡°Ë(compliance-by-default), ±¹¹Î ´ë»ó ¼³¸í°¡´É ÀÚµ¿ÀÀ´äÀÇ °íµµÈ­¸¦½ÇÇà °æ·Î·Î ±¸Ã¼È­ÇÑ´Ù. º» ³í¹®Àº ¹ý·ü°¡ÀÇ ÇØ¼®¡¤ÆÇ´ÜÀ» ´ëüÇϱ⺸´Ù µ¥ÀÌÅÍ¡¤Ç¥ÁØ¡¤Äڵ忡 ±â¹ÝÇÑ ±¸Á¶È­¡¤Å½»ö¡¤°ËÁõÀ» °íµµÈ­ÇÔÀ¸·Î½á, ¹ýÇÐÀÇ Áõ°Å±â¹Ý Àüȯ°ú ±¹¹ÎÀÇ ¹ý·ü Á¢±Ù¼º Á¦°í, ±×¸®°í Åõ¸íÇÏ°í ¿¹Ãø °¡´ÉÇÑ ¹ýü°è·ÎÀÇ ÀÌÇà¿¡ ±â¿©ÇϰíÀÚ ÇÑ´Ù.

This study examines how methodologies from computational social science can be transplanted into law, articulating the concept, historical trajectory, and overseas developments of Computational Law while diagnosing institutional and technical conditions in Korea to propose a Korean-style implementation framework. Using the standard pipelinedata collection, cleaning, exploration/visualization, modeling/inference, and evaluation/reproducibilityas an analytic scaffold, the paper considers how statutes, case law, and administrative rules can be rendered machine-readable and empirically analyzed through statistics, machine learning, and network methods. Internationally, jurimetrics (since the late 1940s) and contemporary initiatives such as Stanford CodeX and the MIT Computational Law Lab demonstrate progress in rule formalization, contract automation, and precedent prediction. In Korea, despite promising resources (e.g., open legislative APIs and large-scale litigation records), bottlenecks persist in the scope/format of judgment disclosure, standardization and machine-readability, and regulatory/ethical uncertainty. To address these constraints, the paper proposes: (i) textualization of judgments and a pilot case-level metadata standard; (ii) Rules as Code pilots in domains with numeric/eligibility criteria, verified by golden-case test suites; (iii) neuro-symbolic integration of LLMs with rule engines and knowledge graphs to strengthen explainability and evidentiary grounding; (iv) reuse governance encompassing de-identification, purpose limitation, access control, and licensing, coupled with a tripartite structure of guidelinessandboxreview committee; and (v) joint education, clinics, and consortia across law, computer science, and statistics/industrial engineering. A phased roadmap (012 months; 13 years; 35 years) details knowledge-graph integration, compliance-by-default modules, and citizen-facing explainable legal assistants. Rather than substituting legal interpretation, the contribution is to systematize structuring, exploration, and verification through data, standards, and code, thereby advancing evidence-based legal scholarship, public access to law, and a more transparent and predictable legal system.



Ű¿öµå¿­±â/´Ý±â ¹öư

, , , , ,

Computational Law, Computational Social Science, Legal NLP and Reasoning, , Data Governance, Korea Roadmap.