MADCS: A Middleware for Anomaly Detection and Content Sharing for Blockchain-Based Systems

Autor(es)

Cardoso e Silva, Alef Vinicius ; Giuntini, Felipe Taliar ; Ranieri, Caetano Mazzoni ; Meneguette, Rodolfo Ipolito ; Garcia, Rodrigo Dutra ; Ramachandran, Gowri Sankar ; Krishnamachari, Bhaskar ; Ueyama, Jo

• Tipo de documento: Artigo Científico
• Fonte: Journal of Network and Systems Management; v. 31, n. 3, p. 29-pg., 2023-07-01.

Resumo

The massive growth in data generation, experienced throughout the current century, has enabled the design of data-driven solutions for various applications. On the other hand, privacy concerns have been raised, especially considering the problems that the leakage of personal data can cause. To address privacy and security issues when dealing with sensitive content, works in the literature have focused on improving protocols for content sharing, primarily by endowing them with anomaly detection modules. However, in Blockchain-based systems, the aggregation of anomaly detection modules to middleware environments is still an under-explored research direction. This paper introduces the Middleware for Anomaly Detection and Content Sharing (MADCS), a new middleware based on a layered structure composed of the application, preprocessing, data analysis and business layers, besides the Blockchain platform. For validation, we built a synthetic dataset of medical prescriptions following an international standard and applied a clustering-based technique for anomaly detection. Experiments demonstrated 85% precision and 78% accuracy in identifying abnormalities in the content-sharing process. The results show that a Blockchain combined with MADCS may contribute to a safer content-sharing network environment. (AU)

• Processo FAPESP: 21/10921-2 - Processamento de imagens para detecção e predição de enchentes
• Beneficiário: Caetano Mazzoni Ranieri
• Modalidade de apoio: Bolsas no Brasil - Pós-Doutorado
• Processo FAPESP: 13/07375-0 - CeMEAI - Centro de Ciências Matemáticas Aplicadas à Indústria
• Beneficiário: Francisco Louzada Neto
• Modalidade de apoio: Auxílio à Pesquisa - Centros de Pesquisa, Inovação e Difusão - CEPIDs
• Processo FAPESP: 18/17335-9 - Explorando DLTs e a inteligência computacional em IoT
• Beneficiário: Jó Ueyama
• Modalidade de apoio: Auxílio à Pesquisa - Regular