[DSRP Evidence](https://dsrpevidence.org/)

# Research on the causal mechanism of prefabricated building accidents: a comprehensive framework integrating association rule mining and graph neural network

Liu et al., 2026, Engineering, Construction and Architectural Management — Engineering

Patterns: [Systems](https://dsrpevidence.org/pattern/systems), [Relationships](https://dsrpevidence.org/pattern/relationships)

## In short

The finding is incomplete without both a relational reading -- causal chains and bridging factors that propagate risk from one accident factor to another -- and a systems reading, since the causal network's whole-level small-world structure is not reducible to any single factor.

## What they found (results)

Analyzing 267 accident cases with association-rule mining and a graph neural network, the study finds the causal network exhibits small-world, heavy-tailed structure, with root and direct causes accounting for about 75% of the most influential nodes and a small number of recurrent causal chains driving propagation.

These researchers were not testing DSRP. The finding is theirs; the correspondence to DSRP is drawn by this site.

[Source](https://doi.org/10.1108/ecam-06-2025-0950)