# Towards a context-dependent numerical data quality evaluation framework

**Authors:** Milen S. Marev, Ernesto Compatangelo, Wamberto Vasconcelos  
**arXiv:** [1810.09399](https://arxiv.org/abs/1810.09399)  
**Date:** 2018-10-22  
**Summary page:** https://www.summarizepaper.com/en/paper/1810.09399v1/

## Key points

- The paper focuses on assessing numeric data quality in the context of Oil and Gas exploration and production workflows.
- The authors highlight the distinct characteristics of numeric data, such as varying significance, unstructured format, mass volume, and real-time processing.
- A novel valuation framework is proposed to address the challenges associated with assessing numeric data quality.
- The framework considers eight relevant data quality dimensions and provides a simple metric for evaluating dataset quality along each dimension.
- The authors argue that their proposed dimensions and metrics effectively capture common quality issues in numerical data.
- This framework is part of a broader research effort to develop an AI-based approach for improving numerical data quality in Oil and Gas workflows.
- The approach aims to enable better decision-making processes by providing a comprehensive assessment of dataset quality using relevant dimensions and metrics.
- Overall, the paper presents a context-dependent numerical data quality evaluation framework specifically tailored for Oil and Gas workflows.

## Summary

This paper titled "Towards a context-dependent numerical data quality evaluation framework" focuses on the assessment of numeric data quality in the context of Oil and Gas exploration and production workflows. The authors emphasize the distinct characteristics of numeric data such as varying significance, unstructured format, mass volume, and real-time processing. To address the challenges associated with assessing numeric data quality, the authors propose a novel valuation framework that takes into account eight relevant data quality dimensions. This framework provides a simple metric to evaluate dataset quality along each dimension. The authors argue that their proposed set of dimensions and corresponding metrics effectively capture the unique quality antipatterns commonly found in numerical data. The introduction of this framework is part of a broader research effort aimed at developing an articulated numerical data quality improvement approach for Oil and Gas exploration and production workflows based on artificial intelligence techniques. This approach enables better decision-making processes in this domain by providing a comprehensive assessment of dataset quality using relevant dimensions and metrics. Overall, this paper contributes to the field by presenting a context-dependent numerical data quality evaluation framework specifically tailored for Oil and Gas exploration and production workflows. The proposed framework offers an effective way to assess dataset quality which can be used to inform decision-making processes in this domain.