Tara Salman, Raj Jain, Lav Gupta, "A Reputation Management Framework for Knowledge-Based and Probabilistic Blockchains," IEEE 1st International Workshop on Advances in Artificial Intelligence for Blockchain (AIChain 2019), held in conjunction with the 2019 IEEE International Conference on Blockchain, Atlanta, July 14, 2019


Recently, leading research communities have been investigating the use of blockchains for Artificial Intelligence (AI) applications, where multiple participants, or agents, collaborate to make consensus decisions. To achieve this, the data in the blockchain storage have to be transformed into blockchain knowledge. We refer to these types of blockchains as knowledgebased blockchains. Knowledge-based blockchains are potentially useful in building efficient risk assessment applications. An earlier work introduced probabilistic blockchain which facilitates knowledge-based blockchains. This paper proposes an extension for the probabilistic blockchain concept. The design of a reputation management framework, suitable for such blockchains, is proposed. The framework has been developed to suit the requirements of a wide range of applications. In particular, we apply it to the detection of malicious nodes and reduce their effect on the probabilistic blockchainsí consensus process. We evaluate the framework by comparing it to a baseline using several adversarial strategies. Further, we analyze the collaborative decisions with and without the malicious node detection. Both results show a sustainable performance, where the proposed work outperforms others and achieves excellent results.

Eprint in Adobe Acrobat format.

Back to the List of Papers
Back to Raj Jain's home page