Data Currency Quality Assessment Based on Multi-sensor

In industry, data quality is crucial for data analysis. The credibility of data analysis relies on the low-quality sensor data in multi-sensor scenarios. Current methods assess data currency quality for each sensor separately, lacking a holistic multi-sensor analysis. To this end, this paper proposes a data quality assessment method based on association currency (DQA-AC). DQA-AC first establishes the relationship between the sampling times of sensors by a directed delay graph. Then it uses the specially designed currency score function and currency fusion algorithm to evaluate the associated currency score of each record. Finally, it computes the data currency quality of sensors based on the results from previous processes. We performed experiments to compare DQA-AC with a baseline, which demonstrates its efficacy and capacity to reflect the influence of sampling delays on data currency quality results. Additionally, we explored the effect of influence factors on data currency quality assessments. Results indicate that with increasing influence factors, results of DQA-AC tend to stabilize.