A unified model of data uncertainty and data relation uncertainty

Data uncertainty (DU), one of the most interesting topics in information science, inevitably leads to data relation uncertainty (DRU). Moreover, DU greatly affects the results of information processing problems such as sorting and top-k queries with uncertain data. However, studies have not clarified the mechanism, mode, and degree of impact of DU on DRU. Hence, we established a unified mathematical model to reveal the precise relation between DU and DRU. The experimental results showed this model's high accuracy under various conditions and its relative error and absolute error of less than 0.2%. This model can be used to quantitatively estimate DRU according to a given DU and vice versa. Unexpectedly, the experiments showed that under given conditions, high DU will not necessarily lead to high DRU. This characteristic is highly significant to the design of data sorting and query algorithms with uncertain data.& COPY; 2023 Elsevier B.V. All rights reserved.

Automated summary

In this study, a unified mathematical model was established to reveal the precise relation between data uncertainty (DU) and data relation uncertainty (DRU). The experimental results showed the high accuracy of this model under various conditions, with a relative error and absolute error of less than 0.2%. The model can be used to quantitatively estimate DRU according to a given DU and vice versa. Unexpectedly, the experiments showed that high DU will not necessarily lead to high DRU, which is highly significant for the design of data sorting and query algorithms with uncertain data.