When combining telematics data with infrastructure and road network data, new opportunities for spatial road safety analysis are opened, this may although introduce challenges. This study proposes an improved method for aggregating smartphonebased telematics data into entities, such as nodes and edges, to support more robust spatial road safety analysis. A telematics dataset comprising multiple trips represented by geospatial coordinates was aggregated to edges using a nearest-edge approach and to nodes using a buffer-based method, introducing inconsistencies. A refined node aggregation method was introduced, restricting data points to edges directly connected to the buffer’s origin node, resulting in node-level variability reductions of approximately 7–50% across features. Principal component analysis and autoencoder were used to analyze the reconstruction of the latent representations for both datasets. The reconstruction performance of these models showed that the proposed approach produces a more complex dataset, reflected in higher autoencoder reconstruction error and a more evenly distributed explained variance in principal components. Global Moran’s I was used to compare spatial autocorrelation between the two different approaches. The proposed method showed lower values, in some cases nearly halved, suggesting reduced artificial similarity and more distinct local variation. Finally, a random forest classifier was used to examine the relationship between telematics features and crash occurrences, assessing how well each dataset aligns with observed crash risk. The proposed approach suggests a cleaner and less redundant representation of the underlying relationships, yielding a clearer crash risk signal.
| ID | pj301 |
| DOI | |
| Tags |
