Declarative Graph Query Language support for Web and Blockchain Decentralized Applications, Analytics, and Compliance

点击翻译按钮获取中文摘要

Blockchain enabled smart contracts and decentralized applications (dApps) provide a level of trust without intermediation and have the potential to benefit industries, government, and society. Blockchains, or Distributed Ledger Technologies (DLTs), consists of participants in a network that store ledger replicas and come to consensus on the validity of a sequence of blocks (of cryptographically signed transactions) chained by hash pointers. As DLTs domains diversify, more complex dApps transacting across multiple blockchains (public and private) are emerging. The objectives of our research are to develop graph-based query languages techniques and tools to address the data management challenges emerging from increasingly complex decentralized applications over public and private blockchains, and across diverse domains. We address three complementary research themes. Theme 1) We will investigate how graph query languages extended with blockchain specific constructs can bring benefits to blockchain dApps. Our previous experience defining and implementing graph traversal languages suggests that query languages with carefully designed expressiveness can be efficiently implemented, supporting optimization and scalability techniques. The richness of the information present within and around blockchains fits well with graph models (e.g., the participants form a network, and the transactions inputs, outputs, and other execution properties define complex information graphs). Theme 2) In addition to improving the support for transactional queries in scripts, our research will look at the opportunities for supporting analytics in the context of DLTs. Supporting a variety of analytic, visualization, and machine learning techniques in DLTs will require pipelines extracting, validating, and transforming data from multiple blockchains and data stores. Theme 3) We plan to build upon our recent work in query-based compliance over privacy logs to develop new techniques to address privacy, compliance, accountability, and enforcement issues in DLTs. Our approach will formalize these new problems and develop new language proposals, attempting to characterize the theoretical properties that are relevant in practical applications. We will implement prototypes and contribute to open source development efforts. We will validate our proposals experimentally, developing novel benchmarks when needed. Our experiments will use real data (public blockchains are a rich source). Our research will provide graph-based query solutions to address key blockchain data challenges, aiming to facilitate the development and analysis of scalable, data rich dApps. We expect that our research results and our open source contributions will have a significant impact on blockchain technology providers and adopters.