CASIA OpenIR  > 复杂系统管理与控制国家重点实验室  > 复杂系统研究
基于区块链的智能合约理论方法与应用研究
王帅
Subtype博士
Thesis Advisor王飞跃
2019-11-29
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Discipline模式识别与智能系统
Keyword区块链 智能合约 预测市场 知识图谱 去中心化自治组织
Abstract

基于区块链的智能合约被称作“区块链2.0”,其不仅具备传统区块链的一般特征,如去中心化、去信任、可追溯、极难篡改等,还兼具了可编程性和一定的自治性,目前已引起政府部门、金融机构、科技企业的高度关注。然而,智能合约虽已呈现出技术和产业创新驱动的发展态势,相关学术研究却严重滞后、亟待跟进。为此,本论文以“智能合约的规范化——智能合约的自动化——智能合约的智能化”为逻辑主线,对智能合约的理论方法与应用展开系统性研究,以期规范其学术内涵、指导其产业实践、启发其未来发展。论文主要工作和贡献如下:

  1. 为实现智能合约的“规范化”,首次提出基于区块链的智能合约六层基础架构模型。给出智能合约的学术定义,并结合当前主流智能合约开发平台,提出智能合约通用运行机制。接着进一步结合其设计流程、构成要素等,将智能合约的全生命周期划分为协商、开发、部署、运维、学习和自毁等六个阶段。在此基础上,提出智能合约六层基础架构模型,模型自底向上依次为基础设施层、合约层、运维层、智能层、表现层和应用层,详述了每一层的组成要素。所提模型既囊括了智能合约全生命周期中的关键技术,又体现出智能合约未来的研究方向与发展趋势。
  2. 在“规范化”的基础上,结合预测市场给出一个智能合约典型案例实现。预测市场具有良好的预测能力,但长期以来也面临着市场易被操纵、安全性差等问题。考虑到预测市场是一种买卖双方把对未来事件的预测以合约的形式进行交易的市场,提出其适合与区块链相结合,尤其是采用智能合约的方式加以实现。提出一种基于智能合约的预测市场合约设计和算法实现方案,并在企业内部私有链上部署了该预测市场。应用实例表明,基于智能合约的预测市场不仅可以实现其预测功能,还克服了以往预测市场存在的市场易被操纵、安全性差等问题。
  3. 为实现智能合约的“自动化”,以领域知识图谱构建为研究和应用背景,提出一种基于智能合约的去中心化知识图谱构建方案并将图谱应用于深度推荐系统。针对传统知识图谱构建方法普遍存在的费时耗力、易被操纵、数据被中心化机构所垄断等问题,提出可以通过众包汇集群体智慧来加以实现,区块链上的智能合约则用于执行众包业务逻辑,使得知识图谱构建过程去中心化、公开透明、可审计且无法篡改。给出了众包投票及其智能合约实现方案。在此基础上,提出一个链上——链下相结合的深度推荐系统,目的是完成工作任务的候选人推荐,助力企业实现业务流程自动化。详述了推荐系统整体架构、知识图谱嵌入方法以及链下深度推荐系统网络模型。最后通过案例研究证明了系统的有效性。
  4. 针对当前智能合约既不“智能”也非“合约”的问题,提出面向智能合约的“智能组件”概念。“智能组件”是一种融合人工智能算法的、可插拔的、自主自治的区块链系统核心功能模块,可以针对特定场景和计算任务自动选择和配置最优组合,本质上它与人工智能的发展史以及编程语言的演进史是一脉相承的,因此其有望是智能合约迈向“智能化”的有效形式与可能实现途径。提出一种基于ASKE (Application Specific Knowledge Engines)的智能预言机实现方案,以此作为所提“智能组件”的一个初步实现。最后,对智能合约/智能组件未来的应用及表现形态——去中心化自治组织(DAO)进行了深入阐释,并提出一个DAO 五层分析框架。
Other Abstract

Blockchain-based smart contracts are known as “Blockchain 2.0”, which not only have the general characteristics of traditional blockchains, such as decentralization, de-trust, traceability, tamper-resistant, etc., but also have programmability and certain autonomy. So they have attracted the attention of government departments, financial institutions, science-and-technology enterprises. However, although smart contracts have shown a development trend driven by technology and industry innovation, relevant academic research is seriously lagging behind and needs to be followed up. To this end, this thesis takes“the standardization of smart contracts — the automation of smart contracts — the intelligence of smart contracts”as the main line of logic, and systematically studies the theoretical methods and applications of smart contracts, in order to standardize their academic connotations, guide their industrial practices, and inspire their future developments. The main work and contributions of the thesis are as follows:

  1. In order to implement the“standardization”of smart contracts, a six-layer basic architecture model of blockchain-based smart contracts is first proposed. The definition of smart contracts is given, and combined with the current mainstream development platforms, the general operating mechanism of smart contracts is proposed. Then, combined with the design process and components, the whole life cycle of smart contracts is divided into six stages: negotiation, development, deployment, maintenance, learning and self-destruction. On that basis, a basic architecture model of smart contracts which employs a six-layer framework, i.e. infrastructures layer, contracts layer, operations layer, intelligence layer, manifestations layer, and applications layer is proposed, and the components of each layer are described in detail. The proposed model not only covers the key technologies in the whole life cycle of smart contracts, but also reflects their future research directions and development trends.
  2. On the basis of“standardization”, and combined with the prediction markets, a typical case realization of smart contracts is given. Prediction markets have good predictive ability, but for a long time they have also faced problems such as easy market manipulation and poor security. Considering that the prediction market is a market in which buyers and sellers trade the forecast of future events in the form of contracts, we think that it is well suited to be combined with the blockchain, especially by means of smart contracts. So we propose a prediction market contracts design and algorithms implementation scheme based on smart contracts, and deploy the prediction market on an internal private blockchain of an enterprise. The application example shows that the smart contracts-based prediction market can not only achieve its forecasting function, but also overcome the problems of easy market manipulation and poor security in the previous prediction markets.
  3. In order to realize the “automation” of smart contracts, we take the domain knowledge graphs construction as the research and application background, and propose a novel decentralized knowledge graphs construction method based on smart contracts, and apply the constructed knowledge graph to the deep recommender system. Aiming at the problems of traditional construction approaches of knowledge graphs, such as laborious and time-consuming, vulnerable to manipulation, and the data is monopolized by the centralized institutions, we propose that it can be realized through crowdsourcing-enabled group intelligence, and blockchain-powered smart contracts are used to execute the business logics of crowdsourcing, thus making the construction process decentralized, transparent, auditable and tamper-resistant. The crowdsourcing voting and its smart contracts implementation scheme are detailed. Then, a deep recommender system consisting of on-chain and off-chain part is proposed, which is designed to complete the candidates recommendation of the work task and help the enterprise to realize the business process automation. The overall architecture of the recommender system, the knowledge graph embedding methods, and the network model of the off-chain deep recommender system are introduced. Finally, the effectiveness of the system is validated by case study.
  4. In view of the fact that the current smart contracts are neither “smart” nor“contract”, the concept of smart contracts-oriented “Intelligent ware” is proposed.“Intelligent ware”is a pluggable, autonomous blockchain system core function module that combines artificial intelligence algorithms. It can automatically select and configure the optimal wares combination for specific scenarios and computational tasks. In essence, “Intelligent ware”comes down in one continuous line with the development history of artificial intelligence and the evolution of programming languages, so it is expected to be an effective form and possible implementation path for smart contracts to move toward “intelligence”. A smart Oracle implementation scheme based on ASKE (Application Specific Knowledge Engines) is proposed as a preliminary implementation of the proposed “Intelligent ware”. Finally, the future applications and manifestations of smart contracts/intelligent wares — the Decentralized Autonomous Organization (DAO) is systematically elucidated, and a reference model for DAO which employs a five-layer architecture is proposed.
Pages136
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/28338
Collection复杂系统管理与控制国家重点实验室_复杂系统研究
Recommended Citation
GB/T 7714
王帅. 基于区块链的智能合约理论方法与应用研究[D]. 北京. 中国科学院大学,2019.
Files in This Item:
File Name/Size DocType Version Access License
基于区块链的智能合约理论方法与应用研究.(8635KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王帅]'s Articles
Baidu academic
Similar articles in Baidu academic
[王帅]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王帅]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.