DeepFilter: A Transformer-style Framework for Accurate and Efficient Process Monitoring

TL;DR

DeepFilter is a novel transformer-based framework designed to improve the accuracy and efficiency of process monitoring by revising the self-attention mechanism, serving as a practical baseline for practitioners.

Contribution

The paper introduces DeepFilter, a revised self-attention mechanism that enhances process monitoring accuracy and efficiency, addressing limitations of existing transformer methods.

Findings

Improved accuracy in process monitoring tasks

Enhanced computational efficiency

Provides a versatile baseline for practitioners

Abstract

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.