Data Analytics on Blockchains

In recent years, blockchains have been exploited in areas way beyond finance, enabling numerous innovative usage scenarios and applications. However, the extension of the existing systems and applications in order to support data persistence on a blockchain is time-consuming. Therefore, this paper proposes a model-driven based approach leveraging smart contracts with the goal to automate data persistence on blockchains. The approach is evaluated in data analytics use cases. According to our results, the proposed approach fully automates the data import and export processes without negatively affecting the predictive power of the models trained using data coming from the blockchain.