Probabilistic model extracts causes from aviation safety reports
Semantic Causal-Factor Inference from Aviation Incident Narratives Using A Variational Autoencoder with Cosine-Similarity-Based Reconstruction
Artificial Intelligence
Summary
Aviation accident reports contain long stories about what happened, making it hard to find the actual causes quickly. The authors created a computer method that learns how to link these long reports to the likely causes experts found. Their method turns the written reports and causes into numbers, then uses a special kind of AI model called a variational autoencoder to guess the causes from the reports. The model showed good accuracy and helped highlight important cause-related words in the reports. This tool could help experts find safety issues faster without replacing their final judgment.
What this means in practice
- •For aviation safety analysts: Automatically identify likely causes in investigation reports to speed up preliminary safety assessments.
- •For transportation data teams: Integrate causal factor extraction into safety data pipelines to improve monitoring of incident trends.
Authors
AZIIDA NANYONGA1, HASSAN WASSWA, UGUR TURHAN, KEITH FRANCIS JOINER, GRAHAM WILD
Abstract
Aviation accident and incident investigations generate extensive unstructured textual information containing evidence relevant to the causes and contributing factors of safety occurrences. Automatically extracting such information is challenging because causal evidence may be distributed across long and complex investigation narratives. This study proposes a semantic causal-factor inference framework combining natural language processing with a variational autoencoder (VAE) to learn the relationship between aviation investigation narratives and expert-reported probable causes. Investigation narratives and their corresponding probable causes are transformed into numerical representations, after which the encoder maps narrative representations to a probabilistic latent space. The decoder estimates representations of the corresponding probable causes and is trained using an objective that combines Kullback-Leibler divergence with cosine-similarity-based semantic reconstruction. The framework was evaluated using 20,919 finalized U.S. National Transportation Safety Board investigation reports from 2005 to 2020. On the held-out test set, the predicted and expert-reported probable-cause representations achieved a mean cosine similarity of 0.786 (SD = 0.120). The predicted representations also yielded interpretable terms associated with causal information in the reports. The results demonstrate the potential of probabilistic latent representation learning for AI-assisted extraction of causal information from aviation safety narratives while retaining expert investigation as the basis for formal causal determination.