Learning-Based Approaches to Predictive Monitoring with Conformal Statistical Guarantees

Cairoli, Francesca; Bortolussi, Luca; Paoletti, Nicola

doi:10.1007/978-3-031-44267-4_26

Cited by 2 publications

(1 citation statement)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Researchers began exploring the integration of machine learning techniques into runtime veriőcation frameworks to enable more adaptive and intelligent monitoring systems [17,32,51]. These approaches aim to detect subtle anomalies and deviations in system behavior, thereby improving the overall accuracy and effectiveness of runtime veriőcation.…”

Section: Integration Of Machine Learningmentioning

confidence: 99%

Practical runtime verification of cross-organizational smart contracts

Kangogo,

Péter,

Klenik

et al. 2024

Preprint

View full text Add to dashboard Cite

Smart contracts in blockchains often implement critical functionality; consequently, their development time software fault detection, removal, and avoidance have been the target of intensive research. However, the applicability of runtime dependability mechanisms to smart contracts has not been explored systematically, mainly due to the transaction pricing models of public blockchains. In this paper, we propose a novel and practical runtime verification approach for Hyperledger Fabric, the leading cross-organizational blockchain platform, as an error detection facility for fault tolerance mechanisms; building on the proposition that, in contrast to the public setting, cross-organizational blockchains have ample computational headroom for their systematic use. We introduce a smart contract entity framework architecture with reusable domain models and encapsulated runtime verification. We describe Hypernate, our open-source implementation of the approach for Java smart contracts, which uses the Java Modelling Language for requirement specification and integrates the OpenJML verification tool. We demonstrate and evaluate the approach on a Fabric implementation of the TPC-C benchmark.

show abstract

Section: Integration Of Machine Learningmentioning

confidence: 99%