Research on Data Error Detection Based on Graph-Based Model

Le résumé fourni par la source

With the advent of the era of big data, all walks of life have emerged and accumulated massive amounts of data. Especially as structured data that supports the operation of various software systems, the success or failure of its data governance methods and results will directly affect the correctness of the conclusions obtained from further data mining and data analysis. As a key part of data governance, data error detection has attracted extensive attention from academia and industry in recent years. At present, the vast majority of error detection methods have good results on text, image and audio data. However, due to the characteristics of structured data such as data sparsity and the lack of prior knowledge of the data set structure, structured data error detection faces considerable challenges. This paper trains a graph-based model to fully capture the data characteristics of unstructured data and conducts application experiments in actual business scenarios. The experimental results show that this model has a higher accuracy rate in structured data error detection compared with several other error detection algorithms and has excellent application prospects in the field of data governance.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.