计算法学方法的运用规则研究

Research on Application Rules of Computational Jurisprudence Methods

计算法学是利用计算机数据处理分析方法在特定主题的法律大数据中寻求变量之间隐性规律的法学研究方法。计算法学本质上是一种实证研究方法,数据、算法和专家组成计算法学方法的三大核心要素。在运用计算法学方法的过程中,研究者可对比不同方法收集数据的特点,选择获取途径,并从真实性、有效性、一般性三个维度考察原始数据质量。综合数据变量类型、研究目标、可选择分析模型等因素构建开放性问题。选用何种计算模型依赖于研究者对法律数据和模型技术的熟识程度,计算模型分析过程是排除研究者干预的过程,对于模型分析的结果,须从模型指标、模型效果、内容价值三个方面检验评估,以期调整参数,得到最优研究结论。

Computational jurisprudence is a legal research method that uses computer data processing and analysis methods to seek implicit laws between variables in legal big data of a specific subject. Computational jurisprudence is essentially an empirical research method, with data, algorithms and experts as its three core elements. In using the method of computational jurisprudence, researchers can compare the characteristics of data collections, choose the acquisition method, and examine the quality of the original data from the three dimensions of authenticity, validity and generality. Openended questions can be constructed by integrating factors such as data variable types,research objectives, and optional analytical models. The selection of calculation model depends on the researcher’s familiarity with legal data and model technology. The calculation model analysis is a process of eliminating researchers’ intervention. As for the results of model analysis, the test and evaluation should be made from model indicators, model effects, and content value in order to adjust the parameters and obtain the optimal research conclusion.