大数据证据真实性审查规则的建构

The Construction of Rules for the Authenticity of Big Data Evidence

囿于海量基础数据的混乱性与算法的专业性、复杂性,大数据证据的真实性审查难以适用传统证据的审查规则,而处于无规则可循的窘境,亟需建构符合其特性的真实性审查规则,以实现其证明价值。具体来讲,作为基础的海量数据在“数量”上需满足采集全量性规则,确保分析挖掘出的潜在信息、规律的正确性和稳定性。同时,在“质量”上需满足大数据整体真实性规则,避免虚假数据产生虚假结果,减损大数据证据的真实性;作为分析工具的算法模型应具备准确性、适配性与可解释性,满足法律与技术双重面向的科学性要求,为大数据证据的真实性提供支撑;从印证的普遍适用性、数据经验的强客观性以及间接证据定罪的规范要求来看,建构分析结果的可印证性规则存在充足理由。通过引入故事模型理论,可发现分析结果的可印证性规则包括内部面向的基础数据印证与外部面向的分析结果印证,其中前者为其他规则提供保障,后者契合大数据证据的证明逻辑。

Due to the chaotic nature of massive basic data and the professionalism and complexity of algorithms,it is difficult to apply the review rules of traditional evidence for the authenticity of big data evidence,and it is in the dilemma of lacking the relevant rules,and it is urgent to construct the authenticity review rules in line with its characteristics in order to realise its evidential value.Specifically,as a foundation,massive amounts of data must meet the rules of completeness in terms of“quantity”to ensure the correctness and stability of potential information and patterns mined from analysis.At the same time,in terms of“quality,”it is necessary to meet the overall authenticity rules of big data to avoid false results from false data and reduce the authenticity of big data evidence.As an analytical tool,algorithmic models should have accuracy,adaptability,and interpretability to meet the scientific requirements of both legal and technical aspects and provide support for the authenticity of big data evidence.In terms of universal applicability of verification,strong objectivity of data experience,and normative requirements for indirect evidence conviction,there is sufficient reason to construct verifiability rules for analysis results.By introducing story model theory,it can be found that the verifiability rules for analysis results include internal-oriented basic data verification and external-oriented analysis result verification.The former provides protection for other rules while the latter fits the logic of big data evidence.