School of Law, Tsinghua University, Beijing 100084, China.
Comput Math Methods Med. 2022 Jul 18;2022:1331237. doi: 10.1155/2022/1331237. eCollection 2022.
Smart contracts are widely employed in many industries as a result of the high-quality development of science and economic technology, as well as the introduction of blockchain, which can automatically conduct retrieval, verification, and payment tasks. Smart contracts as an emerging topic, particularly the study of smart legal contracts, must remain forward-looking, and the smart contract sector cannot wait for the legal status of smart contracts to be resolved before advancing. The relative lag of the law becomes unavoidable due to the unassembled and unpredictable character of the law and thus its legislation. In this paper, we explore the incorporation of smart contracts into the scope of legal regulation, the construction of a series of systems for smart contracts, and the prognosis of smart contracts in terms of contract logic, arbitration process, and formal verification from the current law. Furthermore, a smart contract payment template based on semantic-aware graph neural networks is proposed to address the traditional smart contract vulnerability detection payment template method's low detection accuracy and high false alarm rate, as well as the neural network-based method's insufficient mining of bytecode-level smart contract features. Experiments comparing the method described in this research to comparable methods reveal that the strategy proposed in this study improves all types of indicators significantly.
智能合约在科学和经济技术的高质量发展以及区块链的引入下，在许多行业中得到了广泛应用，可以自动执行检索、验证和支付任务。智能合约作为一个新兴话题，特别是智能法律合约的研究，必须具有前瞻性，智能合约领域不能等到智能合约的法律地位得到解决后才向前推进。由于法律的非组合性和不可预测性及其立法，法律的相对滞后是不可避免的。在本文中，我们从现行法律的角度探讨了将智能合约纳入法律监管范围、构建智能合约的一系列系统，以及智能合约在合同逻辑、仲裁过程和形式验证方面的预测。此外，还提出了一种基于语义感知图神经网络的智能合约支付模板，以解决传统智能合约漏洞检测支付模板方法检测精度低、误报率高以及基于神经网络的方法对字节码级智能合约特征挖掘不足的问题。将本文提出的方法与可比方法进行的实验比较表明，本研究提出的策略显著提高了所有类型的指标。
