An ETC Missed Transaction Data Restoration Approach for Expressways Considering Multiple Scenario Features

ETC data mining is a prominent research area in intelligent expressways, where ensuring high-quality data is crucial. However, the unique characteristics of ETC data pose challenges for existing generic algorithms in data restoration. To overcome this, the paper proposes an approach for restoring missed transaction data, considering multiple scenario features. The focus is on restoring the transaction time and designing targeted strategies for different driving scenarios. The effectiveness of the proposed algorithm is validated using a real ETC dataset and compared with classical machine learning-based algorithms. Experimental results demonstrate the superior restoration accuracy of the proposed approach, achieving a time restoration accuracy (MAE) of 18 s. This outperforms classical machine learning algorithms by at least two-fold, significantly improving the quality of ETC data. The proposed approach meets the timeliness requirements for big data mining and analysis in the transportation domain.