Extracting recurrent scenarios from narrative texts using a Bayesian network: application to serious occupational accidents with movement disturbance

Construction accident prevention: A systematic review of machine learning approaches

BACKGROUND

The construction industry is an important productive sector worldwide. However, the industry is also responsible for high numbers of work-related accidents, which highlights the necessity for improving safety management on construction sites. In parallel, technological applications such as machine learning (ML) are used in many productive sectors, including construction, and have proved significant in process optimizations and decision-making. Thus, advanced studies are required to comprehend the best way of using this technology to enhance construction site safety.

OBJECTIVE

This research developed a systematic literature review using ten scientific databases to retrieve relevant publications and fill the knowledge gaps regarding ML applications in construction accident prevention.

METHODS

This study examined 73 scientific articles through bibliometric research and descriptive analysis.

RESULTS

The results showed the publications timeline and the most recurrent journals, authors, institutions, and countries-regions. In addition, the review discovered information about the developed models, such as the research goals, the ML methods used, and the data features. The research findings revealed that USA and China are the leading countries regarding publications. Also, Support Vector Machine - SVM was the most used ML method. Furthermore, most models used textual data as a source, generally related to inspection reports and accident narratives. The data approach was usually related to facts before an accident (proactive data).

CONCLUSION

The review highlighted improvement proposals for future works and provided insights into the application of ML in construction safety management.

Identifying low-quality patterns in accident reports from textual data

Accident investigation reports provide useful knowledge to support companies to propose preventive and mitigative measures. However, the information presented in accident report databases is normally large, complex, filled with errors and has missing and/or redundant data. In this article, we propose text mining and natural language processing techniques to investigate low-quality accident reports. We adopted machine learning (ML) to detect and investigate inconsistencies on accident reports. The methodology was applied to 626 documents collected from an actual hydroelectric power company. The initial ML performances indicated data divergences and concerns related to the report structure. Then, the accident database was restructured to a more proper form confirming the supposition about the quality of the reports investigated. The proposed approach can be used as a diagnostic tool to improve the design of accident investigation reports to provide a more useful source of knowledge to support decisions in the safety context.

Identifying environmental and health threats in unconventional oil and gas violations: evidence from Pennsylvania compliance reports

With unconventional oil and gas booming in commercial development, its inevitable environmental damage has aroused the public's vigilance. To support the regulation improvement and early-warning system building, it is of great need to learn the regular patterns in recurrent violations both for practitioners and governments. In this respect, we utilized the "Oil and Gas Compliance Report" from the Pennsylvania Department of Environmental Protection from 2000 to 2019, a total of 5737 violation records, to dig out the historical violation patterns. Through LDA (Latent Dirichlet Allocation) analysis combined with the decision tree model, our research attained the following conclusions: first, we find that the LDA themes of violations as "Erosion and sediment" and "Water pollution" are critical factors for "Failed" enforcement results. Therefore, policymakers and practitioners should pay more attention to those two types of accidents. Second, it is noted that counties are also one of the essential features that matter the enforcement results. Third, we need to consider the role of economic punishment dialectically, while it is not a significant feature for successful enforcement results. That is to say, a monetary penalty may not necessarily improve the effectiveness of the company's measurements.

Classification and pattern extraction of incidents: a deep learning-based approach

Classifying or predicting occupational incidents using both structured and unstructured (text) data are an unexplored area of research. Unstructured texts, i.e., incident narratives are often unutilized or underutilized. Besides the explicit information, there exist a large amount of hidden information present in a dataset, which cannot be explored by the traditional machine learning (ML) algorithms. There is a scarcity of studies that reveal the use of deep neural networks (DNNs) in the domain of incident prediction, and its parameter optimization for achieving better prediction power. To address these issues, initially, key terms are extracted from the unstructured texts using LDA-based topic modeling. Then, these key terms are added with the predictor categories to form the feature vector, which is further processed for noise reduction and fed to the adaptive moment estimation (ADAM)-based DNN (i.e., ADNN) for classification, as ADAM is superior to GD, SGD, and RMSProp. To evaluate the effectiveness of our proposed method, a comparative study has been conducted using some state-of-the-arts on five benchmark datasets. Moreover, a case study of an integrated steel plant in India has been demonstrated for the validation of the proposed model. Experimental results reveal that ADNN produces superior performance than others in terms of accuracy. Therefore, the present study offers a robust methodological guide that enables us to handle the issues of unstructured data and hidden information for developing a predictive model.

Police narrative reports: Do they provide end-users with the data they need to help prevent bicycle crashes?

Bicycles gained significant popularity among Americans in 2020. Greater investment in adequate bicycle safety facilities will be needed. Crash data from police will undoubtedly play a role in decision-making. This research evaluated the data quality of text narratives in police reports on bicycle crashes. The aims were to identify situations in which police officers wrote more detail in the narrative text, investigate if longer reports translate to more in-depth crash descriptions, examine the extent to which narrative texts cover details useful for those charged with bicycle safety. This is a 4-year retrospective cohort study of vehicle-vs-bicycle crashes that occurred between January 1, 2009, and December 31, 2012, in Boston, Massachusetts (USA). Police reports were matched with the Pedestrian and Bicycle Crash Analysis Tool (PBCAT) to measure how much information was captured and when reports were more likely to capture more information. Police reports only captured most information in one area of the standardized form (Crash Typing), with average total missingness of over 75%. Longer reports did reduce the amount of missingness, and officers were more likely to write longer reports when they were on the crash site, when there was an injury, when the crash involved an extended car door, and during the day. A 100% increase in the report's words was associated with a four-percentage point reduction in PBCAT missingness. While longer reports result in less missingness when measured against the standardized crash form, the average report still misses most of the information that the form would capture. We recommend that police departments adopt a standardized form to facilitate information capture at the scene of bicycle-vehicle crashes.

Analysis method for revealing human and organisational factors of occupational accidents with movement disturbance (OAMDs)

Slips, trips and other movement disturbances account for 20 to 30% of recorded occupational accidents (OAs). The causal representations of these accidents hamper their prevention. An analysis method dedicated to occupational accidents with movement disturbance (OAMDs) has been developed to change these representations. In France, the causal tree method (CTM) is very commonly used for analysing OAs. An initial version of an OAMD analysis method, which overcomes the problems encountered when analysing these accidents using the CTM, has been developed. This OAMD analysis method was reviewed by six targeted prevention officers and as a result some proposals have been discarded and this initial version has been transformed into three additional CTM modules. The purpose of these modules is to identify human and organisational factors and provide a formal representation of damage caused, beyond bodily injuries. Practitioner summary: A method for analysing occupational accidents triggered by a slip, a trip or any other movement disturbance has been developed in consideration of the practices and constraints in companies. In particular, this method allows us to highlight the human and organisational factors involved in the accident situation. Abbreviations: OA: occupational accident; OAMD: occupational accident with movement disturbance; CTM: causal tree method.

A C-BiLSTM Approach to Classify Construction Accident Reports

The construction sector is widely recognized as having the most hazardous working environment among the various business sectors, and many research studies have focused on injury prevention strategies for use on construction sites. The risk-based theory emphasizes the analysis of accident causes extracted from accident reports to understand, predict, and prevent the occurrence of construction accidents. The first step in the analysis is to classify the incidents from a massive number of reports into different cause categories, a task which is usually performed on a manual basis by domain experts. The research described in this paper proposes a convolutional bidirectional long short-term memory (C-BiLSTM)-based method to automatically classify construction accident reports. The proposed approach was applied on a dataset of construction accident narratives obtained from the Occupational Safety and Health Administration website, and the results indicate that this model performs better than some of the classic machine learning models commonly used in classification tasks, including support vector machine (SVM), naïve Bayes (NB), and logistic regression (LR). The results of this study can help safety managers to develop risk management strategies.

Multiple patent network analysis for identifying safety technology convergence

Purpose

Using the large database of patent, the purpose of this paper is to structure a technology convergence network using various patent network analysis for integrating different results according to network characteristics.

Design/methodology/approach

The patent co-class analysis and the patent citation analysis are applied to discover core safety fields and technology, respectively. In specific, three types of network analysis, which are centrality analysis, association rule mining analysis and brokerage network analysis, are applied to measure the individual, synergy and group intensity.

Findings

The core safety fields derived from three types of network analysis used by different nature of data algorithms are compared with each other to understand distinctive meaning of cores of patent class such as medical safety, working safety and vehicle safety, differentiating network structure. Also, to be specific, the authors find the detailed technology contained in the core patent class using patent citation network analysis.

Practical implications

The results provide meaningful implications to various stakeholders in organization: safety management, safety engineering and safety policy. The multiple patent network enables safety manager to identify core safety convergence fields and safety engineers to develop new safety technology. Also, in the view of technology convergence, the strategy of safety policy can be expanded to collaboration and open innovation.

Originality/value

This is the initial study on applying various network analysis algorithms based on patent data (class and citation) for safety management. Through comparison among network analysis techniques, the different results are identified and the collective decision making on finding core of safety technology convergence is supported. The decision maker can obtain the various perspectives of tracing technology convergence.

State of science: occupational slips, trips and falls on the same level

Occupational slips, trips and falls on the same level (STFL) result in substantial injuries worldwide. This paper summarises the state of science regarding STFL, outlining relevant aspects of epidemiology, biomechanics, psychophysics, tribology, organisational influences and injury prevention. This review reaffirms that STFL remain a major cause of workplace injury and STFL prevention is a complex problem, requiring multi-disciplinary, multi-faceted approaches. Despite progress in recent decades in understanding the mechanisms involved in STFL, especially slipping, research leading to evidence-based prevention practices remains insufficient, given the problem scale. It is concluded that there is a pressing need to develop better fall prevention strategies using systems approaches conceptualising and addressing the factors involved in STFL, with considerations of the full range of factors and their interactions. There is also an urgent need for field trials of various fall prevention strategies to assess the effectiveness of different intervention components and their interactions. Practitioner Summary: Work-related slipping, tripping and falls on the same level are a major source of occupational injury. The causes are broadly understood, although more attention is needed from a systems perspective. Research has shown preventative action to be effective, but further studies are required to understand which aspects are most beneficial.

Harnessing information from injury narratives in the 'big data' era: understanding and applying machine learning for injury surveillance

OBJECTIVE

Vast amounts of injury narratives are collected daily and are available electronically in real time and have great potential for use in injury surveillance and evaluation. Machine learning algorithms have been developed to assist in identifying cases and classifying mechanisms leading to injury in a much timelier manner than is possible when relying on manual coding of narratives. The aim of this paper is to describe the background, growth, value, challenges and future directions of machine learning as applied to injury surveillance.

METHODS

This paper reviews key aspects of machine learning using injury narratives, providing a case study to demonstrate an application to an established human-machine learning approach.

RESULTS

The range of applications and utility of narrative text has increased greatly with advancements in computing techniques over time. Practical and feasible methods exist for semiautomatic classification of injury narratives which are accurate, efficient and meaningful. The human-machine learning approach described in the case study achieved high sensitivity and PPV and reduced the need for human coding to less than a third of cases in one large occupational injury database.

CONCLUSIONS

The last 20 years have seen a dramatic change in the potential for technological advancements in injury surveillance. Machine learning of 'big injury narrative data' opens up many possibilities for expanded sources of data which can provide more comprehensive, ongoing and timely surveillance to inform future injury prevention policy and practice.

Machine learning approaches to analysing textual injury surveillance data: a systematic review

OBJECTIVE

To synthesise recent research on the use of machine learning approaches to mining textual injury surveillance data.

DESIGN

Systematic review.

DATA SOURCES

The electronic databases which were searched included PubMed, Cinahl, Medline, Google Scholar, and Proquest. The bibliography of all relevant articles was examined and associated articles were identified using a snowballing technique.

SELECTION CRITERIA

For inclusion, articles were required to meet the following criteria: (a) used a health-related database, (b) focused on injury-related cases, AND used machine learning approaches to analyse textual data.

METHODS

The papers identified through the search were screened resulting in 16 papers selected for review. Articles were reviewed to describe the databases and methodology used, the strength and limitations of different techniques, and quality assurance approaches used. Due to heterogeneity between studies meta-analysis was not performed.

RESULTS

Occupational injuries were the focus of half of the machine learning studies and the most common methods described were Bayesian probability or Bayesian network based methods to either predict injury categories or extract common injury scenarios. Models were evaluated through either comparison with gold standard data or content expert evaluation or statistical measures of quality. Machine learning was found to provide high precision and accuracy when predicting a small number of categories, was valuable for visualisation of injury patterns and prediction of future outcomes. However, difficulties related to generalizability, source data quality, complexity of models and integration of content and technical knowledge were discussed.

CONCLUSIONS

The use of narrative text for injury surveillance has grown in popularity, complexity and quality over recent years. With advances in data mining techniques, increased capacity for analysis of large databases, and involvement of computer scientists in the injury prevention field, along with more comprehensive use and description of quality assurance methods in text mining approaches, it is likely that we will see a continued growth and advancement in knowledge of text mining in the injury field.

Injury narrative text classification using factorization model

Narrative text is a useful way of identifying injury circumstances from the routine emergency department data collections. Automatically classifying narratives based on machine learning techniques is a promising technique, which can consequently reduce the tedious manual classification process. Existing works focus on using Naive Bayes which does not always offer the best performance. This paper proposes the Matrix Factorization approaches along with a learning enhancement process for this task. The results are compared with the performance of various other classification approaches. The impact on the classification results from the parameters setting during the classification of a medical text dataset is discussed. With the selection of right dimension k, Non Negative Matrix Factorization-model method achieves 10 CV accuracy of 0.93.