Analysis of Railway Accidents' Narratives Using Deep Learning

Abstract

Automatic understanding of domain specific texts in order to extract useful relationships for later use is a nontrivial task. One such relationship would be between railroad accidents' causes and their correspondent descriptions in reports. From 2001 to 2016 rail accidents in the U.S. cost more than $4.6B. Railroads involved in accidents are required to submit an accident report to the Federal Railroad Administration (FRA). These reports contain a variety of fixed field entries including primary cause of the accidents (a coded variable with 389 values) as well as a narrative field which is a short text description of the accident. Although these narratives provide more information than a fixed field entry, the terminologies used in these reports are not easy to understand by a non-expert reader. Therefore, providing an assisting method to fill in the primary cause from such domain specific texts (narratives) would help to label the accidents with more accuracy. Another important question for transportation safety is whether the reported accident cause is consistent with narrative description. To address these questions, we applied deep learning methods together with powerful word embeddings such as Word2Vec and GloVe to classify accident cause values for the primary cause field using the text in the narratives. The results show that such approaches can both accurately classify accident causes based on report narratives and find important inconsistencies in accident reporting 1 .

Classification of User Comment Using Word2vec and Deep Learning

“…In [6] the authors use data from a short text description of railroad accidents in the US. The text descriptions used in this dataset are difficult for non-experts to understand, so the main objective of this study is to find an effective method to fill in the root causes of accidents that will help to give label more accurately.…”

Systematic Literature Review on Data-Driven Models for Predictive Maintenance of Railway Track: Implications in Geotechnical Engineering

“…The RNN unit takes the current and previous input data into consideration at the same time, enabling the method to perform better at predicting future trends given some historical sequence of data [43,48]. Heidarysafa [49] employed an RNN to discover accident causes from the narrative field in the Federal Railroad Administration (USA) reports. The term-frequency and Word2Vec method (where each word is mapped to a vector) is adopted to change the accident text report into sequence vectors.…”

<strong>Cooking is All About People: Comment Classification on Cookery Channels Using Bert and Classification Models (Malayalam-English Mix-Code)</strong>

“…Heidarysafa et al [14] Federal Railroad Administration (FRA) reports. The results indicate that if this process is automated by applying Deep Learning for text analysis, the resultant models can exploit accident narratives which can prove to be useful for safety engineers.…”

“…Analysis of Railway Accidents' Narratives using Deep LearningThis paper[14] states, from 2001 to 2016, rail accidents in the U.S. cost more than 4.6 Billion Dollars. The accident details reported to the Federal Railroad Administration (FRA) consists of entries such as primary cause and a short description.…”