Research on Alignment between Emergency Plan Text and Emergency Disposal Process Mode

Emergency plans are a type of ubiquitous document resource in emergency systems and are easily understood by process executors and non-professionals. By contrast, emergency disposal process models as model-based representations excel at expressing complex logical relation and are suitable for process modelers and professionals. Inconsistency and conflicting description information between emergency plans and emergency disposal process model can lead to different process executions and expectations. Aligning emergency plans to emergency disposal process models is a primary step to maintain their consistency. An automatic alignment approach between emergency plan and emergency disposal process models is proposed in this paper. Firstly, process information is extracted respectively from emergency plans and emergency disposal process models based on dependency parsing of Language Technology Platform (LTP). Secondly, similarity between nodes of process models and sentences of emergency plans is computed based on a uniform feature representation. Thirdly, given five constraints, alignment is established based on similarity between nodes and sentences. Finally, alignment results are further optimized. In addition, experimental evaluation is introduced and it is illustrated that the proposed approach can be used to assist experts to align emergency plans according to emergency disposal process models.