# LMT: A Bayesian Framework for Causal Discovery from Textual Alarm Records in Manufacturing Systems

arXiv: https://arxiv.org/abs/2606.09892 · HTML: https://mlsanity.com/paper/2606.09892/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 53 references that are in the map.
- First posted on arXiv 3 Jun 2026.

## Builds on (most cited first, 22 of 22)

- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [OpenAI GPT-5 System Card](https://mlsanity.com/paper/2601.03267/) (2025)
- [DAGs with NO TEARS: Continuous Optimization for Structure Learning](https://mlsanity.com/paper/1803.01422/) (2018)
- [Causal Reasoning and Large Language Models: Opening a New Frontier for Causality](https://mlsanity.com/paper/2305.00050/) (2023)
- [DiBS: Differentiable Bayesian Structure Learning](https://mlsanity.com/paper/2105.11839/) (2021)
- [Learning Granger Causality for Hawkes Processes](https://mlsanity.com/paper/1602.04511/) (2016)
- [Causal Discovery with Language Models as Imperfect Experts](https://mlsanity.com/paper/2307.02390/) (2023)
- [Causal Discovery from Temporal Data: An Overview and New Perspectives](https://mlsanity.com/paper/2303.10112/) (2023)
- [EventGPT: Event Stream Understanding with Multimodal Large Language Models](https://mlsanity.com/paper/2412.00832/) (2024)
- [THP: Topological Hawkes Processes for Learning Causal Structure on Event Sequences](https://mlsanity.com/paper/2105.10884/) (2021)
- [LearnDA: Learnable Knowledge-Guided Data Augmentation for Event Causality Identification](https://mlsanity.com/paper/2106.01649/) (2021)
- [Learning Hawkes Processes from a Handful of Events](https://mlsanity.com/paper/1911.00292/) (2019)
- [Mitigating Hallucinations in Large Language Models via Causal Reasoning](https://mlsanity.com/paper/2508.12495/) (2025)
- [Visual Causality Analysis of Event Sequence Data](https://mlsanity.com/paper/2009.00219/) (2020)
- [LLM Cannot Discover Causality, and Should Be Restricted to Non-Decisional Support in Causal Discovery](https://mlsanity.com/paper/2506.00844/) (2025)
- [A Survey of Event Causality Identification: Taxonomy, Challenges, Assessment, and Prospects](https://mlsanity.com/paper/2411.10371/) (2024)
- [Event Extraction in Large Language Model](https://mlsanity.com/paper/2512.19537/) (2025)
- [TNPAR: Topological Neural Poisson Auto-Regressive Model for Learning Granger Causal Structure from Event Sequences](https://mlsanity.com/paper/2306.14114/) (2023)
- [Uncovering Bias Paths with LLM-guided Causal Discovery: An Active Learning and Dynamic Scoring Approach](https://mlsanity.com/paper/2506.12227/) (2025)
- [Synthesis of Temporal Causality](https://mlsanity.com/paper/2405.10912/) (2024)
- [Causality-informed Anomaly Detection in Partially Observable Sensor Networks: Moving beyond Correlations](https://mlsanity.com/paper/2507.09742/) (2025)
- [A Predictive Model Based on Transformer with Statistical Feature Embedding in Manufacturing Sensor Dataset](https://mlsanity.com/paper/2407.06682/) (2024)

## Abstract (arXiv)

Textual event records, such as alarm logs, have become an increasingly common data source in engineering and manufacturing systems. Beyond identifying correlations or recurring patterns, engineers are often interested in understanding which types of events causally trigger or influence other events during system operation. Textual event descriptions may contain semantic clues about such causal relationships, and recent large language models (LLMs) provide a promising tool for extracting these signals. However, relying solely on LLM-encoded textual information is insufficient for accurate causal discovery, since semantic patterns do not directly reveal causal mechanisms and may confuse causation with correlation or frequent sequential patterns. To address these challenges, we propose LMT, a Bayesian causal discovery framework for engineering event data that jointly leverages textual descriptions and timestamps. Specifically, LMT first uses LLMs to extract semantic causal signals from event descriptions and constructs a prior distribution over causal graphs among event types or event clusters. It then incorporates temporal evidence through a Poisson-process-based likelihood, allowing the LLM-informed prior to be refined by timestamp-based statistical evidence. By integrating the textual and temporal information, LMT produces a causal graph that is both interpretable and data-supported. Simulation studies show that the proposed framework is effective across different settings and is especially advantageous in small-sample alarm-event scenarios.