The process monitoring task is characterized by stringent demands for accuracy and efficiency. Current transformer-based methods, characterized by self-attention for temporal fusion, exhibit limitations in accurately understanding the semantic context and efficiently processing monitoring logs, rendering them inadequate for process monitoring. To address these limitations, we introduce DeepFilter, which revises the self-attention mechanism to improve both accuracy and efficiency. As a straightforward yet versatile approach, DeepFilter provides an instrumental baseline for practitioners in process monitoring, whether initiating new projects or enhancing existing capabilities.
@article{arxiv.2501.01342,
title = {DeepFilter: A Transformer-style Framework for Accurate and Efficient Process Monitoring},
author = {Hao Wang and Zhichao Chen and Licheng Pan and Xiaoyu Jiang and Yichen Song and Qunshan He and Xinggao Liu},
journal= {arXiv preprint arXiv:2501.01342},
year = {2026}
}