【摘要】 跨域立案是最高人民法院推行的一项便民利民的立案措施,其中要求对当事人进行适当的法律释明,具体包含案件适用法律条文、适当引导当事人明确诉请、固定争点、有效举证等。为了探索智能化跨域立案,对案件进行适用法律条文自动推荐,基于民间借贷案件,提出了基于案例推理的推荐方法。将历史案例作为推理的源数据,设计案件指标体系对案例数值化表示,并分析指标的重要程度,通过相似度计算对最相似的案例进行查找,提出一种排序权重方法,对相似案例的适用法律条文进行加权,设置权重阈值作为推荐法律条文的判断标准,并给出适用法律推荐结果。通过收集的民间借贷案由数据实验,结果表明基于案例推理的法律条文推荐具有较好的准确率和召回率,所推荐的适用法律条文对于跨域立案中的法律释明业务具有参考价值。更多还原

【Abstract】 Cross-domain filing is a quite convenient approach for the benefit of the people by the Supreme People’s Court, in which judges are required to perform legal interpretation for the parties, including telling the applicable laws of cases, and guiding the parties to clarify the petition, fix the disputes, and provide effective evidence properly, etc. In order to explore intelligent cross-domain case filing and give recommendation of applicable laws automatically, we propose an applicable law recommendation method based on case-based reasoning(CBR) for private loan cases. With historical cases as the source data for reasoning, an index system is designed to digitize the cases, and the importance of the indicators is analyzed. The most similar cases searching through similarity calculation, a weight-based ranking method is proposed to set weight thresholds as a criterion for recommending legal provisions and weight the applicable legal provisions to generate the results of applicable law. The experiments on the collected data about private lending cases show that the recommendation of legal provisions based on case-based reasoning achieves high accuracy and recall. The recommended applicable laws have a high reference value for legal interpretation in cross-domain filing.更多还原