Data Reliability. You are solely responsible for the accuracy and completeness of all data transmitted by you using the Services.
Data Reliability. The reliability of data is crucial to develop a computa- tional model or to support an empirical claim [29]. Over the years, research has focused on various ways to select reliable samples from unreliable datasets based on peer effects and social networks [30]. In an unreliable data environ- ment, peer-based sample selection has emerged as a promising approach to train models and select high-confidence samples [2]. This method involves using a group of models to collectively judge and select samples to train the model [31]. This technique aims to improve the performance of the model in scenarios that involve unreliable data, while also mitigating the influence of confirmation bias [2]. Peer-based sample selection has the potential to enhance the accuracy and reliability of learning systems in situations where data quality is uncertain [32]. When different models produce consistent re- sults, it indicates that they have a similar understanding of categories and can be expected to perform consistently [32]. However, it is imperative to consider that attaining agreement does not invariably guarantee validity; nevertheless, it is probable that they would agree on reliable samples to a greater extent [32]. Although our work draws inspiration from the aforemen- tioned approaches, our fundamental aim is to address the issue of noisy-label image classification through the application of a novel sample selection.
Data Reliability. Data reliability is a crucial foundation when building data on an interesting topic. The data is sufficiently complete and accurate [14]. It is related to the quality of the captured data. Reliability can also be expressed as the number of failures over a time window [15]. Our goal is to answer the following research questions: RQ1: What implementations of Blockchain are used to support social selling processes in small businesses? How can a food Blockchain traceability platform improve the trust in a social selling scheme? RQ2: Can a smart contract be used to measure the reliability of the data stored in a food Blockchain traceability platform? Could this negatively affect the number of transactions performed on the network per second? We limited the review to those related to the food supply chain or product traceability and security, considering that the following research was selected for the final evaluation. In [16], the authors reviewed different Blockchain applications in food traceability, showing the Blockchain platform used, the traced product, and the use cases. Finally, they discuss the role of this technology in enhancing traceability processes. Authors in [17,18] show how it is possible to solve problems associated with the lack of trust in traditional centralized systems using blockchain technology. However, there is no mention of trust problems related to the reliability of the data stored on the network. Authors in [19] explore the potential of Blockchain as a sustainability tool for food supply chains; they conclude that the technology is used as a tool for sustainability and addressing sustainability challenges. On the other hand, authors like [20,21] propose using Blockchain technology to improve data reliability, transparency, and food safety in an agri-food supply chain. Each research shows multiple benefits this technology provides in traceability, information access, decision making, etc. Authors in [22] offer a solution for untrusted food traceability information using Blockchain technology combined with IoT technologies to collect and update traceability information. Unlike the discussed research, our work proposes smart contracts to validate the data using the previously agreed conditions by the traceability partners. We observe that some authors like [10,11,23–25] have been addressing data reliability in Blockchain networks from different perspectives proposing storage systems that can be incrementally updated to ensure data re...
