Next Article in Journal
Shear Transfer Mechanism between CFRP Grid and EPS Rigid Foam Insulation of Precast Concrete Sandwich Panels
Next Article in Special Issue
Competencies Driving Waste Minimization during the Construction Phase of Buildings
Previous Article in Journal
Physico-Chemical Substantiation of Obtaining an Effective Cement Composite with Ultrafine GGBS Admixture
Previous Article in Special Issue
Assessing Factors Affecting Construction Equipment Productivity Using Structural Equation Modeling
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Public Opinion Mining on Construction Health and Safety: Latent Dirichlet Allocation Approach

1
Civil and Architectural Engineering Institute, Panzhihua University, Panzhihua 617000, China
2
Sustainable Real Estate Research Center, Hong Kong Shue Yan University, Hong Kong, China
3
City 4.0 Lab, Queensland University of Technology, Brisbane, QLD 4000, Australia
4
Rajamangala University of Technology Tawan-Ok, Bangkok 10400, Thailand
*
Author to whom correspondence should be addressed.
Buildings 2023, 13(4), 927; https://doi.org/10.3390/buildings13040927
Submission received: 6 March 2023 / Revised: 26 March 2023 / Accepted: 28 March 2023 / Published: 31 March 2023

Abstract

:
The construction industry has been experiencing many occupational accidents as working on construction sites is dangerous. To reduce the likelihood of accidents, construction companies share the latest construction health and safety news and information on social media. While research studies in recent years have explored the perceptions towards these companies’ social media pages, there are no big data analytic studies conducted on Instagram about construction health and safety. This study aims to consolidate public perceptions of construction health and safety by analyzing Instagram posts. The study adopted a big data analytics approach involving visual, content, user, and sentiment analyses of Instagram posts (n = 17,835). The study adopted the Latent Dirichlet Allocation, a kind of machine learning approach for generative probabilistic topic extraction, and the five most mentioned topics were: (a) training service, (b) team management, (c) training organization, (d) workers’ work and family, and (e) users’ action. Besides, the Jaccard coefficient co-occurrence cluster analysis revealed: (a) the most mentioned collocations were ‘construction safety week’, ‘safety first’, and ‘construction team’, (b) the largest clusters were ‘safety training’, ‘occupational health and safety administration’, and ‘health and safety environment’, (c) the most active users were ‘Parallel Consultancy Ltd.’, ‘Pike Consulting Group’, and ‘Global Training Canada’, and (d) positive sentiment accounted for an overwhelming figure of 85%. The findings inform the industry on public perceptions that help create awareness and develop preventative measures for increased health and safety and decreased incidents.

Graphical Abstract

1. Introduction

Construction workers engage in various activities that put them at risk, such as falling from roofs, being struck by heavy construction equipment, and electrocution [1]. The construction industry is one of the riskiest industries worldwide [2]. The United Nations’ International Labor Organization (ILO) in 2019 recorded 2.8 million worker deaths globally, with more than 6500 people dying every day from work-related illnesses and 1000 from occupational accidents. Workers in the construction business still endure high rates of fatal and non-fatal injuries and accidents despite improvements in construction health and safety, technology, and training [3].
According to [3], despite the construction industry only accounting for 6% of the US labor force, it accounted for around 20% of work-related deaths in the country. For instance, there were 174,100 construction injuries in the US in 2020 [4]. As per the [2], 36 construction workers and 5 members of the general public were killed in the UK on average each year over the past 5 years, and 61,000 workers experienced non-fatal injuries from 2018 to 2021 [2]. In 2022, there were more than 30 casualties in China’s construction sector [5]. Studies conducted in Ethiopia have shown that construction workers’ injury rates ranged from 30% to 84.7% [6].
While an injury on a construction site is not entirely avoidable [7], there is an increase in the usage of the Fourth Industrial Revolution (Industry 4.0) technologies for enhancing construction site safety [8,9]. Thus far, research on construction health and safety via social media includes Sina Weibo [10], Twitter [11], and YouTube [12]. Nevertheless, to our knowledge, no research has investigated construction issues through Instagram, not to mention construction health and safety issues. Therefore, this study sheds light on public perceptions of construction health and safety by analyzing Instagram posts.
The research question is: How does the public perceive construction health and safety on Instagram? The answers to this question inform construction health and safety industry managers to share knowledge via Instagram and improve related measures on sites.
Following this introduction, Section 2 presents the literature review on construction health and safety and Instagram. Section 3 introduces the research methods. Section 4 elaborates on data and sentiment analyses. Section 5 presents the findings and discussions. Section 6 summarizes the theoretical and practical contributions, limitations, future research directions, and concludes the paper.

2. Literature Background

2.1. Construction Health and Safety

Construction works include buildings (e.g., new work, repair, additions, and alterations), civil engineering (e.g., road and railway), and specialized construction activities (e.g., demolition, electrical, plumbing, joinery, plastering, painting, and glazing) [2]. Health and safety issues are key factors affecting the sustainable development of large-scale construction projects [13].
Over the last decade, much research has been carried out on construction health and safety management. For example, the outbreak of COVID-19 has significantly impacted the construction industry. Nnaji et al. [14] surveyed construction companies and worker satisfaction in the United States. The results indicated that workers tend to use technologies such as video conferencing apps and wearable sensing devices to reduce the spread of COVID-19 on sites.
Some scholars discussed the issues in the project environment from workers’ perspectives. For example, [15] took a large construction site in Iran to study site health and safety and formulate strategies to balance risks and costs to reduce health and safety risks during construction. Others researched the associations between company profiles, project characteristics, and safety practices. Gurmu [16] found that health and safety policies, health and safety plans, and hazard analysis are the three most important practices for improving construction project productivity. The level of health and safety practices increased with project costs, company experience, and size. Project delays negatively related to the level of health and safety practices [16].
Others highlighted the impact of gender and age on construction health and safety. Mariam et al. [17] conducted scientometric and meta-analyses of the health and safety of gender in construction. Women’s health and safety research grew slowly. The main hazards faced by women in the construction industry were biologically related [17]. Peng and Chan [18] studied the adjusted working conditions to address the health and health deterioration of older construction workers in Hong Kong. They found that physical and mental health were key factors affecting the occupational safety of older construction workers. An organization’s health and safety culture and level of maturity affect the likelihood of accidents. Williams et al. [19] identified the health and safety culture of the construction industry in developing economies. They used the construction industry in Ghana as an example to assess health and safety maturity.

2.2. Social Media and Public Perceptions

Many countries are now harnessing the power of social media to improve people’s lives. Social media allows interaction with others in many ways [20]. Due to convenience, a considerable amount of information is disseminated through social media [11]. More people share and publish their ideas with increased social media, including public users, company managers, academic scholars, government officers, etc. Companies are increasingly relying on social media to analyze social media activity and professionalize their social media engagement, such as Twitter [21]. Some studies suggest that articles with more social media followers have higher visibility and citations [22].
Social media platforms have become essential for the broader dissemination of research articles. Authors, readers, and publishers host different social media campaigns for articles [22]. Social media can provide construction health and safety educational resources, especially during COVID-19. It improves the public’s awareness and perception of construction health and safety. Different social groups can benefit from social media, where important news, the latest information, and updates can be quickly and efficiently disseminated.
With the help of social media, many construction health and safety companies have quickly begun to promote construction safety products. Zeng and Li [10] studied construction health and safety hazard awareness on the Web of Science and Weibo between 1991 and 2021. They explored the different foci of awareness of construction health and safety between academia and the public. Yao et al. [11] surveyed construction health and safety knowledge-sharing on Twitter and performed a social network analysis. Scholars explored how YouTube facilitates knowledge-sharing in construction health and safety [12]. Previous research found that many YouTube videos mentioned other social media such as “Facebook”, “Instagram”, “Twitter”, and “LinkedIn”. These demonstrated the information-sharing on social networking platforms [12]. However, none of these utilized Instagram data despite its popularity.
Based on the abovementioned literature background, this study identifies a research problem that is scarcely studied—the limited understanding of public opinion on construction health and safety. This research problem led us to explore how the public perceives construction health and safety on Instagram. In order to bridge this knowledge gap, this study was conducted through investigating the following three research questions: (1) What is the publics’ main foci on Instagram about construction health and safety? (2) What kind of users on Instagram are active in the topic of construction health and safety? (3) What are the user sentiments on Instagram about construction health and safety?

3. Research Method

3.1. Instagram Big Data Analytics

Instagram is one of the largest and most influential social media platforms worldwide [23]. It was founded in 2010 by software engineer Michel Krieger and former Googler Kevin Systrom [23]. Since its launch in October 2010, Instagram has rapidly become one of the top four social networks in the world [24]. More than 2 billion people use Instagram every month, growing faster than other social networks such as Facebook and Snapchat [23].
Instagram is currently the fourth most popular social media network in the world by the number of users, and its parent company Facebook owns four of the six top social networks (Facebook, WhatsApp, Messenger, and Instagram) [24]. Almost a quarter of the world’s active internet users visit the app monthly [24,25]. Of the 4.18 billion active mobile internet users, 23.92% visit Instagram every month, equivalent to the combined population of Europe and North America [24]. According to CNBC, Facebook had 2.91 billion monthly active users in October 2021, but its expansion is slower than Instagram’s. Instagram’s user base has doubled, while Facebook’s has only grown by 30% [26].
Scholars have studied construction health and safety on social media. For example, [11] adopted social network analysis to examine Twitter construction health and safety knowledge-sharing. Despite Instagram being one of the main social media platforms offering invaluable big data for researchers to analyze various urban and built environment phenomena [27], studies of construction health and safety on Instagram are scarce. SooHoo and SooHoo [28] examined the importance of educating teenagers on electric scooter safety equipment through content analysis of posts on social platforms such as Instagram. Couteau et al. [29] collected do-it-yourself (DIY) product recipes on social media such as Instagram and analyzed 275 eye makeup DIY recipes and their possible safety issues.

3.2. Sentiment Analysis

In recent decades, social media has become a well-known platform for expressing the feelings of ordinary people around the world. Opinion plays a vital role in uncovering some critical decisions. Sentiment analysis or opinion mining is a method used to mine the thoughts or feelings of the general population [30,31]. Studies have demonstrated and evaluated the ability of various volumetric, affective, and social network approaches to predict critical decisions from online social media platforms [30].
Sentiment analysis and applications for social media have been developed by academia to extract insights and understand public opinions on specific topics to support businesses, manufacturers, government agencies, and policymakers’ decision-making and planning [32]. Social media-based sentiment analysis is generally more representative because it is based on many participants’ opinions, which can enhance reliability [32]. However, sentiment analysis is vital for providing policymakers with adequate mechanisms for understanding customer and citizen attitudes that can be used in decision-making processes and future planning [33].
Research on sentiment analysis in construction health and safety is limited to date. Some scholars have applied sentiment analysis in construction health and safety on Twitter, and the study found that sentiment valence showed a positive correlation between favorites and retweets [11]. Others explored the role of YouTube in promoting construction health and safety knowledge and investigated how emotions affect users’ viewing and commenting behaviors through semantic analysis [12]. With the rise of Weibo, sentiment analysis for short texts has also become a focus of social network analysis. Zeng and Li [10] adopted AI sentiment to analyze the Weibo users’ foci and perceptions of construction health and safety hazards’ awareness. Based on the researchers mentioned above, this paper adopted a sentiment analysis of construction health and safety posts on Instagram.

3.3. Research Methods

This studied used the keywords “construction safety”, or “construction health”, or “construction health and safety” and collected 17,835 posts, 93,707 paragraphs, and 116,975 sentences. The results showed that there had been a steady increase in content from 2015 to 2022. The year 2022 recorded the highest number of construction health and safety content on Instagram (Figure 1).
After obtaining the Instagram posts, this studied utilized KH Coder text-mining software for analysis. It has been widely used in many research fields in Japanese, such as sociology, economics, linguistics, and education [34,35]. KH Coder has functions such as word frequency statistics, part-of-speech analysis, clustering, correlation analysis, and visualization (such as histograms and clustering maps) [36]. This study used KH Coder to perform word frequency statistics, part-of-speech analysis, and clustering analysis of keywords and authors and to obtain the visualization results. It can support text data mining in Japanese, Chinese, English, French, German, Spanish, Portuguese, Italian, Russian, etc. The language on Instagram on this topic was mainly English, and the posts were imported into KH Coder for word selection and pre-processing to ensure the accuracy of the extracted data. The software extracts high-frequency words. Then, through the keyword search function, we observed the context information where the high-frequency keywords were located. Finally, the software generated a keyword co-occurrence relationship diagram, etc., and analyzed the distribution and connection of the main elements.
We used the function of the KH Coder 3 Folder to remove stop words which were not related to our research, such as “W”, and “PRP” in “Parts of Speech”, etc., and some general items were removed, i.e., constructions, construction, health, health and safety, healthy, safe, safer, safest, construction health, and safely. Some other keywords with the same semantic meanings, such as “people, person”, “awareness, Awareness”, “builder, builders, BUILDERS”, “building, buildings, Buildings”, “construction site, Construction Site, construction sites”, “construction worker, Construction Worker, Construction Workers, construction workers”, “fall, falls”, “service, services”, “general contractor, General Contractor, general contractors”, “COVID, COVID-19”, “info, information”, etc., were combined in the second-round keyword analysis.
Tableau is a business intelligence presentation tool positioned for agile development and implementation of data visualization. It provides interactive visual analytics and dashboard applications. When working with large-scale, multidimensional data, it can instantly show the patterns of data presentation from different angles. In this study, keywords and authors’ word frequency data from KH Coder were imported to Tableau, attributes such as shape, color, size, and labels were defined according to the “dimension” and “measure” settings, and the visual analysis results were obtained. The larger the area of the circle, the higher the frequency of the word in the source text [34], then the hot spots and popular authors were obtained.
This study used Simply Sentiment (CX Data Science) to perform sentiment analysis. It transformed unstructured texts into digital data and identified positive and negative sentiments in the text and generated a set of numerical scores based on this analysis. In addition, Simply Sentiment could generate a dashboard with a graphical representation of the overall data processed. For each text item analyzed, Simply Sentiment’s results included a score of total positive sentiment, total negative sentiment, the overall sentiment in the text (positive–negative), and a text label (positive, negative, or neutral) that represents the overall polarity of the text [37].

4. Results

4.1. Instagram Popular Contents and KOL on Construction Health and Safety

The study showed that the first content about construction health and safety appeared in 2015. Salmaraz posted the first content about construction health and safety, “Finally came in the mail! #ConstructionHealth&SafetyTechnician by #BCSP #CHST”. This might be because Instagram users focused on entertainment, celebrities, etc., in the early years.
Since then, there was an increase in Instagram promotion for events such as Construction Safety Week. Based on [38], the organization was initially launched in 2014, when more than 40 national and global construction companies, comprising the Construction Industry Safety Initiative (CISI) Group and the Incident and Injury-Free (IIF) CEO Forum, joined forces with a common goal: to inspire everyone in the industry to become a leader in safety. Through analysis of contents in 2016, they were mainly found with the hashtag of “construction safety week”, which was initially started in 2014, and became a fully integrated annual campaign with a growing number of members and sponsors in 2016 [38]. Especially in 2016, the Construction Industry Federation (CIF) called on construction companies of all sizes to participate in Construction Safety Week [39], and companies started to use Instagram to share knowledge about construction health and safety. Thus, the content about construction health and safety on Instagram increased to 14 in 2016 and has kept increasing. Besides, construction workers were five times more likely to contract COVID-19 than the public from 2019 to 2022 [3]. Thus, the posts on Instagram about construction health and safety increased in 2019. More than 70 national and global construction companies are now coming together to host Safety Week [38]. Construction Safety Week is Instagram’s most influential word in construction health and safety. It was an impressive and popular topic in the construction industry via Instagram.

4.2. Keyword Analysis Results

The foci on Instagram were mainly divided into two perspectives: (a) construction workers, such as team, site, work, job, etc., and (b) construction health and safety training, such as training, construction safety week, occupational health and safety administration (OSHA), courses, etc. We analyzed the word frequencies via Tableau 2021.3. The influential keywords were calculated and are visualized in Figure 2. The keywords ranked with the top ten highest frequencies were: construction safety week (n = 5666), safety first (n = 3806), team (n = 3473), site (n = 3339), work (n = 3102), project (n = 2853), training (n = 2388), construction workers (n = 2372), job (n = 2320), and industry (n = 2129) (Table 1).
The most influential collocation was “Construction Safety Week” (Figure 2 and Table 1). Construction Safety Week is an influential organization in the construction industry worldwide. The organizer used social media and encouraged relative companies to strengthen the construction industry’s safety culture and performance by sharing the best practices, tools, and resources [38]. Thus, the word can obtain the highest frequency on Instagram.
“Safety first” ranked among Instagram’s top two influential words about construction health and safety. Putting safety first is key to helping reduce the high injury rate in the construction industry, and companies that put safety first will save money over time. Everyone benefits from fostering a safety culture on construction sites [3]. Thus, safety first played a vital role in construction health and safety.
“Team” ranked as the third most influential keyword. Construction workers need to pay special attention to teamwork to ensure that the entire operation is smooth, efficient, and most importantly, safe for all involved. Industrial construction projects often consist of numerous professionals. A team can include engineers, managers, architects, investors, and contractors for various applications. Each professional must deal with a specific aspect of the overall project. The team here mainly involves two concepts, teamwork and team building, which are crucial to success. Team building focuses on team formation, and teamwork focuses on aspects of teamwork ability. Understanding this critical difference can help managers play the team leader or field team member more effectively.

4.3. Jaccard Coefficient Co-Occurrence Cluster Analysis

This study utilized the Jaccard coefficient to calculate the strength of co-occurrence, and the strongest co-occurrences are drawn as network edges in Equation (1) below:
JC   ( M 1 ,   M 2 ) = | M 1 M 2 | | M 1 M 2 | = | M 1 M 2 | | M 1 | + | M 2 | | M 1 M 2 |
The intersection M1 ∩ M2 is a multiset which is composed by all elements in M1 and M2 and repeated many times in the lower cardinality multiset. M1 ∪ M2 is a multiset which is composed of all M1 and M2 elements and is repeated many times in the higher cardinality multiset. Assume fM is the multiplicity function (f) of a multiset M in domain {0,1, n−1}, where 0 ≤ i ≤ n−1, and fM(i) shows the frequency that the element i occurs in M.
The minimum term frequency (min. TF) was set as 475, and the number of selected words was 141 when specifying the type of edges to words. Here, 22 clusters were divided by different colors (Figure 3). The largest cluster, with an orange color, included: “safety training”, “OSHA”, “HSE”, “safety tips”, “safety professionals”, “business”, “pike consulting”, “training”, “safety work”, “workplace safety”, and “safety consulting”. It focused on construction health and safety training. This cluster aims to share health and safety training knowledge and education information via Instagram.
The words “learn”, “keep”, “link”, “bio”, “website”, “visit”, “information”, and “please” were in the second cluster, with a pink color. It mainly described the users’ actions to learn the training courses or knowledge. The blue cluster, including: “site”, “job”, “company”, “best”, “practice”, and “continue”, ranked third. It emphasized the site’s safety and management in the construction industry. Apart from these three clusters, the light green cluster, including: “construction safety week”, “safety first”, “safety week”, “workplace safety”, and “PPE”, and the dark green cluster, including: “team”, “project”, and “member”, were also influential, as shown in Figure 3. This refers to the fact that many construction/site safety week organizers share the news of these campaigns via Instagram. On the other hand, another blue cluster contained “site”, “job”, “company”, “best”, “practice”, and “continue”, indicating that some companies might also use Instagram to promote safety internally and share the best practice with their employees.

4.4. Popular Construction Health and Safety Content Analysis

As shown in Table 2, the most and the secondary influential content was with the hashtag #Construction Safety Week, which obtained 4099 and 3399 publish-like counts, respectively. The first one was mainly focused on the training courses live on LinkedIn and posted by user A in 2020. The content was about construction health and safety training in life and encouraged users to join them. The results had some agreement with the results of word frequencies. Successful event hosting became a vital motivation for people to share related information on Instagram.
The secondary content with 3399 publish-like counts was “AUTOSTOP™ Technology Helps Protect Workers”. It was posted by user B in 2022 and developed innovative tools and worksite solutions. The contents of “Learn More About How MX FUEL™ Jobsite Lighting Helps Prevent Trips and Falls by Eliminating Cords” ranked third and had 2719 published numbers. It was also published by the user of C in 2022. These two contents were about some companies’ recommendations for new technology for users. It indicated that users on Instagram liked companies publishing practical and innovative productions or technologies.
As shown in Table 2, these popular contents mainly focused on occupational training courses, new and innovative technologies and products, and health and safety knowledge that benefited workers. Thus, a content analysis may help managers find users’ and market demands, which showed safety in first place rather than commercial profiles.

4.5. User Analysis

As presented in Figure 4 and Table 3, the most influential user was a consultancy company (user count = 282) with over 30 years of residential and commercial building surveying experience (Parallel Consultancy Ltd., London, UK, 2022, https://www.instagram.com/parallelconsultancyltd/ accessed on 29 August 2022). Its official website shows that it is a London-based building surveying firm with the capability and experience to cover London, Surrey, Essex, and surrounding areas (Parallel Consultancy 2022, https://www.parallelconsultancy.co.uk/about-us/ accessed on 29 August 2022). On Instagram, it has 351 followers. The second most active user was a leading provider of safety solutions for those working in hazardous industries, including construction safety, industrial safety, public works, and municipal safety (Pike Consulting Group, Fayetteville, GA, USA, 2022, https://pikegroup.net/ accessed on 29 August 2022). It is a company of licensed security professionals specializing in security management systems. From Pike Consulting Group (2022), Mitchell Smith and Gary DeGeorge Sr., founders of Pike Consulting Group, designed their clients’ security management systems to be OSHA-compliant solutions. Then, Global training Canada (user count = 197) ranked third. It is an internationally recognized company providing technical training courses to various industries for thousands of workers annually on-site, operator safety, and industry-specific certification requirements (Global Group of Companies, Aston Manor, South Africa, 2022, https://globaltrainingcentre.com/serve_content.cfm?Page=Why%20Choose%20Us accessed on 29 August 2022).
All in all, the influential users were mainly companies, but not individuals. Instagram is a platform for companies to advertise their products and services. These companies can benefit workers and the construction industry, and they can also get fame by sharing construction health and safety training courses, ideas, successful projects, knowledge, etc., on Instagram. More details are shown in Table 4.

4.6. Sentiment Analysis

Most public users on social media are emotional. Sentiment analysis plays a vital role in social media. The study adopted CX Data Science to calculate the sentiment scores about construction health and safety on Instagram posts. There were three top scores, including a +’ve score (positive score), -’ve score (negative score), and a neutral score. As shown in Figure 5a, the +’ve score accounted for 70%, the -’ve scored only 8%, and the neutral score accounted for 4%. Among all the kinds of sentiment, positive, negative, and neutral sentiments were 85.4%, 4.7%, and 9.8%, respectively (Figure 5b, Figure 6, Table 5).
In Figure 7, the posted information from Instagram about construction health and safety showed a higher positive sentiment trend, such as “safe”, “working”, “thank you”, “help”, “great”, “support”, “proud”, “thanks”, and “respective”, as shown in the first column of Figure 7a. While from Figure 7c, “injuries”, “injury”, “hard”, “accidents”, “stress”, “issues”, “suicide”, and “effort” were all negative sentiments, but accounted for less than positive sentiments. This indicated that the public’s perception of construction health and safety on Instagram was positive, although construction is one of the most dangerous industries. On the other hand, most users are organizations or companies that share construction health and safety training or products for the public that are beneficial for workers. Some scholars indicated that sentiment analysis on social media understands broader public opinion behind specific topics to support decision-making by businesses, manufacturers, government agencies, and policymakers and planning [32]. Thus, this result can be beneficial for companies to decide how to share the construction health and safety knowledge products, etc., on social media, especially on Instagram.
As for Figure 8, the most frequently mentioned verbs and adjectives were positive, as shown in AI sentiment analyses. The most mentioned positive adjectives included “proper”, “right”, “proud”, “professional”, “available”, “happy”, “free”, “important”, “new”, etc. The highest-frequency adjective—“new”—refers to new construction health and safety coming to new heights. It mainly includes four aspects: (a) new materials and structures, i.e., new mortar, lightweight and comfortable fall protection, steels, (b) new health and safety initiatives and measures, i.e., new health and safety protocols, procedures, standards, policies, and calibration, (c) new techniques and innovations, i.e., new 4D health and safety tools, fabrication, multi-disciplinary facility replacement project, and (d) new industrial facilities and environment, i.e., new workwear, bannisters and flooring, hard-hat, safety equipment, and tools. The second highest-frequency adjective is “important”. It mainly expresses essential policies and protocols, environments, and tools for individuals and teams, i.e., safety on site is essential, the right tools for construction safety jobs are important, PPE and RRB are necessary, companies’ foci on workers’ wellbeing is important, safety is vital in construction because the people on the team are more important than any project, and it is crucial for construction crews to wear the right gear.
As for Figure 9, these words ranked in the top 20 among the higher frequencies of adjectives. For verbs, “work”, “follow”, “get”, “keep”, “learn”, “help”, “provide”, “ensure”, “thank”, “check”, etc., ranked in the top 20 among higher frequencies of verbs and belonged to positive sentiment. The highest-frequency verb was “work”, and work safely, work with a team, work with a positive attitude, flexible work, working professionals, work together, work in conjunction with the finest state-of-the-art equipment, etc., were mainly work-related phrases on Instagram in the context of construction health and safety. Then, the secondary verb was “follow”, which mainly referred to following the construction health and safety training courses, pieces of knowledge, etc.

4.7. Latent Dirichlet Allocation Topic Model

Topic models are mainly used for semantic analysis and text-mining problems in natural language processing. For example, they are utilized to collect, categorize, and reduce the dimensionality of text by topic. The Latent Dirichlet Allocation (LDA) is one of the topic model approaches that can be used to analyze the topic distribution of documents [40,41]. It reveals the topic of each document set in the form of a probability distribution so that after analyzing some documents to extract their topic distribution, it can be subject-clustered or use text classification according to the topic distribution [41]. Each topic is represented with a word distribution. Documents are shown in LDA as random mixtures of latent subjects. It computes a corpus’ probability using the formula:
p ( C | x , b ) = d = 1 m P ( Φ d | x ) ( n = 1 n d T a p ( T a | Φ d ) p ( W d n | T a , b ) ) d Φ d
where C is the corpus of the documents (D), x is each document topic distribution’s Dirichlet previous concentration parameter, while b refers to the corpus-level parameter, Φd represents the document-level variable, n is the number of words in the document, T a is the topic assignment, Wdn refers to the nth word in the dth document, and m represents the documents that needed to be analyzed [42]. The LDA topic model was analyzed from the KH Coder via the Fit a Topic Model. We set the number of topics as ten via the function of topic model (LDA) options. The 10 LDA topic models about construction health and safety on Instagram are shown in Figure 10. The 10 topic models from #1 to #10 can be relatively summarized as training service (#1, 0.726), team management (#2, 0.979), training organization (#3, 0,75), workers’ work and family (#4, 0.858), users’ action (#5, 0.978), site management (#6, 0.897), industry activities (#7, 0.818), industry development (#8, 0.815), training method (#9, 0.764), and responsibility (#10, 0.768). Among them, the topic of team management (#2, 0.979) had the highest score, which includes team (0.250), project (0.204), keep (0.132), jobsite (0.067), member (0.066), part (0.058), support (0.054), crew (0.054), work (0.050), and priority (0.044).

5. Findings and Discussion

Previous research evidenced the effectiveness and importance of social media analytics to capture public opinions and institutional capabilities [43,44]. Based on this capability in this study, the analysis of Instagram posts through the Latent Dirichlet Allocation topic model helped consolidate public opinion on construction health and safety. The analysis of public perceptions on construction health and safety generated insights and helped in addressing the three research questions of this study: (1) What is the publics’ main foci on Instagram about construction health and safety? (2) What kind of users on Instagram are active in the topic of construction health and safety? (3) What are the user sentiments on Instagram about construction health and safety? The key findings are presented and discussed below.

5.1. The Most Popular Topic and Hashtag

There were 5666 “Construction Safety Week” posts which were about construction health and safety on Instagram. It was an essential annual activity in the construction industry. The result showed that successful construction safety event hosting became an essential motivation for people to share related information on Instagram. Applying hashtags about construction health and safety could be beneficial and quickly spread among relevant stakeholders. The other popular topics and/or hashtags were “Safety First” and “Construction Team”. Team management is an important topic, as per LDA topic model analysis.
At the same time, on Instagram, the public preferred to use a hashtag to record their published topics about construction health and safety. The result was different from the former studies in construction health and safety. Some scholars may not focus on other social media, such as research on construction safety and hazard awareness on Sina Weibo [10]. Besides, as social media platforms need to classify and effectively distribute online content, hashtags are essential to classify content and promote the contents to others. Adding hashtags could increase the exposure and discoverability of the social media content and raise the interaction among users who may not even know the content writers. This study found that “Construction Safety Week”, “Safety First”, and “Construction Team” as event tags and construction industry labels attracted many public users. It is recommended that industry companies use these tags on social media to post content on Instagram, to promote the positive dissemination of construction health and safety knowledge. It also has a positive impact on the development of the construction industry.

5.2. The Largest Cluster and the Most Liked Content

The largest cluster mainly included the keywords: “safety training”, “OSHA”, “HSE”, “safety tips”, “safety professionals”, “business”, “pike consulting”, “training”, “safety work”, “workplace safety”, and “safety consulting”. Thus, construction health and safety training was popular among Instagram users. Previous research demonstrated the relationship between information-sharing and social networking platforms, such as “Facebook”, “Instagram”, “Twitter”, and “LinkedIn” [12], however this research did not reveal a direct relationship in the results. The keywords indirectly revealed that similar contents were shared through Instagram. Instagram users share practical construction health and safety training courses or projects, mainly focusing on teamwork or team building and innovative AI technologies, which could be beneficial to construction workers.

5.3. The Most Active Users

The top 3 active users were Parallel Consultancy Ltd. (followers = 351), Pike Consulting Group (followers = 115), and Global Training Canada (followers = 218). They are surveying or safety training companies. Unlike other social media, for example, Sina Weibo, the most popular users are governments in the context of construction safety [10]. To promote the brand, companies adopted social media to share content related to construction safety. Nevertheless, active users may receive little focus on their posts. That is to say, active users’ sharing contents may be less popular. For example, the most liked content (like number = 4099) among all the research samples was from user A (followers = 1.02 million). The safety training, especially on LIVE, is more exciting and valuable for users. Thus, active users may change their strategies to share more construction safety training courses or innovative technologies to benefit them. To enhance post-sharing effectiveness, Instagram writers might focus more on the construction team, site and environment, workers, etc.

5.4. Positive and Negative Sentiments

All in all, the number of positive sentiments was 18 times that of negative sentiments from users’ posts on Instagram about construction health and safety. Although construction is a hazardous industry, people posted more positive information on Instagram. This was especially true among construction or safety training companies, who preferred sharing positive information to help workers to decrease accidents. By adopting training courses and knowledge and promoting and popularizing new tools, materials, structures, facilities, rules, and products on social media, the workers’ perception of construction health and safety will increase. For example, the active user Parallel Consultancy Ltd. (2021) posted: “Funding for the new 4D health and safety tool was provided through a Small Business Research Initiative—LRB-SBRI-RRB—innovation competition funded by HS2 and managed by Innovate UK. The competition aims to deliver innovation that will drive project efficiencies and processes for use during the design and construction phase of HS2.” Chauhan et al. [30] have verified the ability of various sentiments to predict critical decisions from online social media platforms. These companies and organizations can share more positive posts, such as new materials and structures, new health and safety initiatives and measures, new techniques and innovations, and new industrial facilities and environments, to offer information to the public users and workers to avoid injuries.

6. Conclusions

6.1. Theoretical Implications

The construction industry is one of the most critical industries in the country. However, professionals in this industry often face multiple health and safety risks. These include slips and trips, falls, exposure to hazardous products, etc. Construction health and safety are worldwide issues that can be shared on social media such as Instagram. In total, 30.94% of the world’s 4.48 billion active social media users regularly access Instagram [25]. With the increasing number of people on Instagram, we can research the public’s foci and increase companies’ and workers’ perceptions about construction health and safety via big data analysis. This study analyzed the posts about construction health and safety on Instagram Via Python, Tableau, KH Coder, and AI sentiment. The influential and popular foci, active users, and sentiment were analyzed. The results can benefit more companies’ managers and workers in the construction industry. Despite Instagram being one of the main social media platforms offering invaluable big data opportunities for researchers to analyze various urban and built environment phenomena [27], studies of construction health and safety on Instagram are scarce. The research filled the research gap by studying Instagram posts’ content that shared construction health and safety knowledge. Furthermore, the study generated a consolidated understanding on public opinion on construction health and safety.
Social media’s “construction safety”-related big data analysis can improve the public’s perception of building health and safety awareness, improve risk prediction for construction health and safety management, and promote the sustainable development of projects. The first content about construction health and safety appeared in 2015 on Instagram. The most popular phrase about construction health and safety on Instagram is Construction Safety Week, a frequent hashtag. The foci on Instagram were primarily divided into two perspectives: (a) construction workers, such as team, site, work, job, etc., and (b) construction safety training, such as Construction Safety Week, training, OSHA, courses, etc. The active users of Instagram were mainly construction companies, and the top three were Parallel Consultancy Ltd., Pike Consulting Group, and Global Training Canada. They primarily provide consulting services and training courses in construction safety. We may speculate that an increase in profit drives online construction health and safety marketing and promotion activities via Instagram.

6.2. Practical Implications

Analyzing topics in construction health and safety through social media contributes to the decision-making of company managers and the development of safety in the construction field. Hashtags of successful event hosting, such as “Construction Safety Week”, have become an important motivation that motivates people to share and focus on related information on Instagram. It indicated that users on Instagram liked companies’ publishing of practical and innovative productions or technologies. Through the big data analysis, “team” should be taken into consideration by companies to promote the construction safety training. The population from Instagram analyzed here is different from that from social media analyzed by previous scholars, which is conducive to exploring views on construction safety from different dimensions and enriching the portraits of the population from social media.
The sentiment analysis on construction health and safety on Instagram was mainly positive, accounting for 85.4%. Based on previous studies, sentiment analysis is a method used to mine the feelings of the general population [30,44]. It can be concluded that through the dissemination of social media, the field of construction safety is developing in a positive direction, and users’ awareness is also more positive. Therefore, construction companies can reduce the public’s negative perception of construction health and safety awareness through construction safety training. Sentiment analysis and applications for social media can support businesses, manufacturers, government agencies, and policymakers’ decision-making and planning [32]. Thus, companies have adopted Instagram to share some construction health and safety knowledge, products, etc. It would give workers positive public sentiment and benefit their occupational health and safety.

6.3. Research Limitations and Research Directions

While interpreting the study findings, the following limitations should be noted: (a) There is more than one language used on Instagram. According to [24], Instagram is the most popular in India, with 180 million users, followed by the US (170 million), Brazil (110 million), Indonesia (93 million), and Russia (61 million). Still, we only chose English as the primary language to collect data. (b) Some construction workers and managers, such as older workers, may not use social media. Thus, the results might reflect younger generations’ knowledge-sharing activities. (c) According to the Ghost Data report, 9.5%, or 95 million Instagram users, may be fake bots [24], impacting the data collection and analysis. We may further study the same topic based on different countries or economic situations, for example, developed countries compared with developing countries, in the context of construction health and safety on Instagram. Besides, in the future, it would be worth analyzing the share of occupational health and safety applications and tracking the extent to which they are used by various participants in the investment process.

Author Contributions

Conceptualization, L.Z. and H.Z.; methodology, L.Z. and R.Y.M.L.; software, L.Z. and H.Z.; validation, L.Z., R.Y.M.L. and T.Y.; formal analysis, L.Z. and T.Y.; investigation, L.Z.; resources, L.Z. and T.Y.; data curation, L.Z. and H.Z.; writing—original draft preparation, L.Z. and H.Z.; writing—review and editing, L.Z., R.Y.M.L. and T.Y.; visualization, L.Z. and H.Z.; project administration, L.Z. and R.Y.M.L.; funding acquisition, L.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the PhD Starting Research Fund from the Panzhihua University (No. 035200153).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

Publicly available datasets were analyzed in this study. This data can be found here: https://www.linkedin.com.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. OSHA. Construction Industry. Occupational Safety and Health Administration. 2022. Available online: https://www.osha.gov/ (accessed on 26 August 2022).
  2. Health and Safety Executive (HSE). Construction Statistics in Great Britain. 2021. Available online: https://www.hse.gov.uk/statistics/industry/construction.pdf (accessed on 26 March 2023).
  3. BigRentz. 25 Construction Safety Statistics for 2022. 2022. Available online: https://www.bigrentz.com/blog/construction-safety-statistics (accessed on 8 August 2022).
  4. BLS. Employer-Reported Workplace Injuries and Illnesses–2021. Available online: https://www.bls.gov/news.release/pdf/osh.pdf. (accessed on 26 March 2023).
  5. Construction Industry Focus. Exposure|More than 30 Casualties in the Construction Field in 2022! Summary of Recent Typical Security Incidents. 2022. Available online: https://mp.weixin.qq.com/s/upl6ZEXIPW2zFzqu0Kl_7Q (accessed on 26 August 2022).
  6. Meseret, M.; Ehetie, T.; Hailye, G.; Regasa, Z.; Biruk, K. Occupational injury and associated factors among construction workers in Ethiopia: A systematic and meta-analysis. Arch. Environ. Occup. Health 2021, 77, 328–337. [Google Scholar] [CrossRef]
  7. Ranju, W. Top Causes of Global Construction Fatalities, and How to Avoid Site Risks. Construction Week. 2019. Available online: https://www.constructionweekonline.com/people/training/255830-top-10-causes-of-construction-deaths-and-how-to-prevent-site-accidents (accessed on 26 August 2022).
  8. Oh, T.K.; Kwon, Y.J.; Oh, B.-H.; Gwon, Y.-I.; Yoon, H.-K. Suggestions for safety coordinator’s roles at each construction stage (client, designer, supervisor, and contractor) to improve safety and health activities in South Korea. Saf. Sci. 2020, 133, 104994. [Google Scholar] [CrossRef]
  9. Regona, M.; Yigitcanlar, T.; Xia, B.; Li, R.Y.M. Opportunities and Adoption Challenges of AI in the Construction Industry: A PRISMA Review. J. Open Innov. Technol. Mark. Complex. 2022, 8, 45. [Google Scholar] [CrossRef]
  10. Zeng, L.; Li, R.Y.M. Construction safety and health hazard awareness in Web of Science and Weibo between 1991 and 2021. Saf. Sci. 2022, 152, 105790. [Google Scholar] [CrossRef]
  11. Yao, Q.; Li, R.Y.M.; Song, L.; Crabbe, M.J.C. Construction safety knowledge sharing on Twitter: A social network analysis. Saf. Sci. 2021, 143, 105411. [Google Scholar] [CrossRef]
  12. Yao, Q.; Li, R.Y.M.; Song, L. Construction safety knowledge sharing on YouTube from 2007 to 2021: Two-step flow theory and semantic analysis. Saf. Sci. 2022, 153, 105796. [Google Scholar] [CrossRef]
  13. Wu, X.; Chong, H.-Y.; Wang, G.; Li, S. The Influence of Social Capitalism on Construction Safety Behaviors: An Exploratory Megaproject Case Study. Sustainability 2018, 10, 3098. [Google Scholar] [CrossRef] [Green Version]
  14. Nnaji, C.; Jin, Z.; Karakhan, A. Safety and health management response to COVID-19 in the construction industry: A perspective of fieldworkers. Process. Saf. Environ. Prot. 2022, 159, 477–488. [Google Scholar] [CrossRef]
  15. Zamankhani, J.S. Evaluation of Safety and Health Conditions in Large Scale Construction Projects. Eur. J. Sustain. Dev. 2018, 7, 175–181. [Google Scholar] [CrossRef]
  16. Gurmu, A.T. Identifying and prioritizing safety practices affecting construction labour productivity: An empirical study. Int. J. Prod. Perform. Manag. 2019, 68, 1457–1474. [Google Scholar] [CrossRef]
  17. Mariam, A.T.; Olalusi, O.B.; Haupt, T.C. A scientometric review and meta-analysis of the health and safety of women in construction: Structure and research trends. J. Eng. Des. Technol. 2020, 19, 446–466. [Google Scholar] [CrossRef]
  18. Peng, L.; Chan, A.H. Adjusting work conditions to meet the declined health and functional capacity of older construction workers in Hong Kong. Saf. Sci. 2020, 127, 104711. [Google Scholar] [CrossRef]
  19. Williams, J.; Fugar, F.; Adinyira, E. Assessment of health and safety culture maturity in the construction industry in developing economies: A case of Ghanaian construction industry. J. Eng. Des. Technol. 2019, 18, 865–881. [Google Scholar] [CrossRef]
  20. Brinkman, C.S.; Gabriel, S.; Paravati, E. Social achievement goals and social media. Comput. Hum. Behav. 2020, 111, 106427. [Google Scholar] [CrossRef]
  21. Benthaus, J.; Risius, M.; Beck, R. Social media management strategies for organizational impression management and their effect on public perception. J. Strat. Inf. Syst. 2016, 25, 127–139. [Google Scholar] [CrossRef]
  22. Karmakar, M.; Banshal, S.K.; Singh, V.K. Does presence of social media plugins in a journal website result in higher social media attention of its research publications? Scientometrics 2020, 124, 2103–2143. [Google Scholar] [CrossRef]
  23. Mansoor, I. Instagram Revenue and Usage Statistics (2022). Business of Apps. 2022. Available online: https://www.businessofapps.com/data/instagram-statistics/#:~:text=Instagram%20was%20a%20hit%2C%20rapidly,1%20billion%20monthly%20active%20users (accessed on 27 August 2022).
  24. Brian, D. Instagram Demographic Statistics: How Many People Use Instagram in 2022? BACKLINKO. 2022. Available online: https://backlinko.com/instagram-users (accessed on 27 August 2022).
  25. Statista. Global Digital Population as of April 2022. Statist. 2022. Available online: https://www.statista.com/statistics/617136/digital-population-worldwide/ (accessed on 27 August 2022).
  26. Salvador, R. Instagram Surpasses 2 Billion Monthly Users While Powering Through a Year of Turmoil. CNBC. 2021. Available online: https://www.cnbc.com/2021/12/14/instagram-surpasses-2-billion-monthly-users.html (accessed on 27 August 2022).
  27. Yigitcanlar, T.; Kankanamge, N. Urban Analytics with Social Media Data: Foundations, Applications and Platforms; CRC Press: New York, NY, USA, 2022. [Google Scholar]
  28. SooHoo, C.; SooHoo, J. Importance of Educating Teenagers on Appropriate Safety Gear for E-Scooters. Comment on “Follow-Up Investigation on the Promotional Practices of Electric Scooter Companies: Content Analysis of Posts on Instagram and Twitter”. JMIR Public Health Surveill. 2020, 6, e18945. [Google Scholar] [CrossRef]
  29. Couteau, C.; Girard, E.; Coiffard, L. Analysis of 275 DIY recipes for eye cosmetics and their possible safety issues. Int. J. Cosmet. Sci. 2022, 44, 403–413. [Google Scholar] [CrossRef]
  30. Chauhan, P.; Sharma, N.; Sikka, G. The emergence of social media data and sentiment analysis in election prediction. J. Ambient. Intell. Humaniz. Comput. 2020, 12, 2601–2627. [Google Scholar] [CrossRef]
  31. Yigitcanlar, T.; Kankanamge, N.; Preston, A.; Gill, P.S.; Rezayee, M.; Ostadnia, M.; Xia, B.; Ioppolo, G. How can social media analytics assist authorities in pandemic-related policy decisions? Insights from Australian states and territories. Health Inf. Sci. Syst. 2020, 8, 37. [Google Scholar] [CrossRef]
  32. Omar, A.; Hamouda, W.I. A Sentiment Analysis of Egypt’s New Real Estate Registration Law on Facebook. Int. J. Adv. Comput. Sci. Appl. 2021, 12, 656–663. [Google Scholar] [CrossRef]
  33. Regona, M.; Yigitcanlar, T.; Xia, B.; Li, R. Artificial intelligent technologies for the construction industry: How are they perceived and utilized in Australia? J. Open Innov. Technol. Mark. Complex. 2022, 8, 16. [Google Scholar] [CrossRef]
  34. Wang, Z.; Yuan, L.; Zhang, W. A Comparative Analysis of STEM Education Research Models between China and Japan Based on the Text Mining of KH Coder. Mod. Distance Educ. Res. 2020, 32, 56–63. [Google Scholar] [CrossRef]
  35. Cioban, S.; Lazăr, A.R.; Bacter, C.; Hatos, A. Adolescent Deviance and Cyber-Deviance. A Systematic Literature Review. Front. Psychol. 2021, 12, 748006. [Google Scholar] [CrossRef]
  36. Ben Khalfallah, H.; Jelassi, M.; Demongeot, J.; Ben Saoud, N.B. Decision support systems in healthcare: Systematic review, meta-analysis and prediction, with example of COVID-19. AIMS Bioeng. 2023, 10, 27–52. [Google Scholar] [CrossRef]
  37. CX Data Science. Sentiment Analysis for Microsoft Excel. CX Data Science. 2023. Available online: https://www.cxdatascience.com/ (accessed on 1 February 2022).
  38. Construction Safety Week. Safety Week|We Are Stronger & Safer Together. 2022. Available online: https://www.constructionsafetyweek.com/about-us/who-we-are/ (accessed on 29 August 2022).
  39. Construction Safety Week. 07 Sep Construction Safety Week 2016. Construction Safety Week. 2016. Available online: https://niso.ie/construction-safety-week-2016/ (accessed on 1 September 2022).
  40. Makita, K.; Suzuki, H.; Koike, D.; Utsuro, T.; Kawada, Y.; Fukuhara, T. Labeling Blog Posts with Wikipedia Entries through LDA-Based Topic Modeling of Wikipedia. J. Internet Technol. 2013, 14, 297–306. [Google Scholar] [CrossRef]
  41. Garg, M.; Rangra, P. Bibliometric Analysis of Latent Dirichlet Allocation. DESIDOC J. Libr. Inf. Technol. 2022, 42, 105–113. [Google Scholar] [CrossRef]
  42. Eligüzel, N. Analyzing society anti-vaccination attitudes towards COVID-19: Combining latent dirichlet allocation and fuzzy association rule mining with a fuzzy cognitive map. Fuzzy Optim. Decis. Mak. 2023, 22, 1–28. [Google Scholar] [CrossRef]
  43. Kankanamge, N.; Yigitcanlar, T.; Goonetilleke, A. How engaging are disaster management related social media channels? The case of Australian state emergency organisations. Int. J. Disaster Risk Reduct. 2020, 48, 101571. [Google Scholar] [CrossRef]
  44. Yigitcanlar, T.; Kankanamge, N.; Regona, M.; Ruiz Maldonado, A.; Rowan, B.; Ryu, A.; Li, R.Y.M. Artificial intel-ligence technologies and related urban planning and development concepts: How are they perceived and utilized in Australia? J. Open Innov. Technol. Mark. Complex. 2020, 6, 187. [Google Scholar] [CrossRef]
Figure 1. Instagram posts on construction health and safety by year.
Figure 1. Instagram posts on construction health and safety by year.
Buildings 13 00927 g001
Figure 2. The influential keywords with high frequencies in construction health and safety on Instagram.
Figure 2. The influential keywords with high frequencies in construction health and safety on Instagram.
Buildings 13 00927 g002
Figure 3. Co-occurrence network of words on construction health and safety on Instagram.
Figure 3. Co-occurrence network of words on construction health and safety on Instagram.
Buildings 13 00927 g003
Figure 4. Influential users who mentioned construction health and safety on Instagram.
Figure 4. Influential users who mentioned construction health and safety on Instagram.
Buildings 13 00927 g004
Figure 5. Total items processed and net sentiment. Note: Pos means positive, Neu means neutral, and Neg means negative.
Figure 5. Total items processed and net sentiment. Note: Pos means positive, Neu means neutral, and Neg means negative.
Buildings 13 00927 g005
Figure 6. Average number of words per item of text and percentage containing a URL or email address.
Figure 6. Average number of words per item of text and percentage containing a URL or email address.
Buildings 13 00927 g006
Figure 7. Tag clouds: (a) all sentiment, (b) positive sentiment, and (c) negative sentiment.
Figure 7. Tag clouds: (a) all sentiment, (b) positive sentiment, and (c) negative sentiment.
Buildings 13 00927 g007
Figure 8. Frequencies of adjectives about the sentiment analysis on construction health and safety.
Figure 8. Frequencies of adjectives about the sentiment analysis on construction health and safety.
Buildings 13 00927 g008
Figure 9. Frequencies of verbs about the sentiment analysis in construction health and safety.
Figure 9. Frequencies of verbs about the sentiment analysis in construction health and safety.
Buildings 13 00927 g009
Figure 10. LDA topic model about construction health and safety on Instagram.
Figure 10. LDA topic model about construction health and safety on Instagram.
Buildings 13 00927 g010
Table 1. The top 20 keywords with higher term frequencies in construction health and safety on Instagram.
Table 1. The top 20 keywords with higher term frequencies in construction health and safety on Instagram.
No.WordsTF (Term Frequencies)
1Construction safety week5666
2Safety first3806
3Team3473
4Site3339
5Work3102
6Project2853
7Training2388
8Construction workers2372
9Job2320
10Industry2129
11Construction sites2103
12Construction life2094
13OSHA2052
14Courses2003
15Worker1827
16Construction industry1714
17Link1710
18Contractors1652
19New1609
20Employee1530
Table 2. Influential content with more likes in IG construction health and safety posts.
Table 2. Influential content with more likes in IG construction health and safety posts.
Publish UserAccount TypePublish TimePublish Content Publish-Like Counts
Asafety training companies/safety product service company15 September 2020 04:21:16It’s #ConstructionSafetyWeek, and we’re getting tough on safety. Join us LIVE on LinkedIn this Wednesday, 9/16 at 12:30 pm ET for a training session.4099
B 4 May 2022 06:00:17Safer Handheld Coring. Learn More About How AUTOSTOP™ Technology Helps Protect Workers Core So They Can Core with More Confidence with The MX FUEL™ Handheld Core Drill. Link in Bio.
#ConstructionSafetyWeek #MilwaukeeTool #NothingButHeavyDuty
3399
Csafety training companies/safety product service company5 May 2022 06:00:12Trips And Falls Are the Leading Cause of Fatal Workplace Injuries. Learn More About How MX FUEL™ Jobsite Lighting Helps Prevent Trips and Falls by Eliminating Cords.
Photo Credit: @completelycordless
#ConstructionSafetyWeek #MilwaukeeTool #NothingButHeavyDuty #LightTheSite
2719
Dsafety training companies/safety product service company4 May 2022 03:18:46The 60V MAX* Cordless Dust Extractor keeps the Jobsite safe, easy and efficient so you can set your mind at ease and focus on the task.2417
E 6 May 2022 06:00:07Milwaukee Tool Provides Universal Cordless Dust Management Solutions That Deliver The Simplest Way To Be OSHA Compliant. Learn More, Link in Bio.
#ConstructionSafetyWeek #MilwaukeeTool #NothingButHeavyDuty
2130
Fsafety training companies/safety product service company3 May 2022 06:00:09Hand Injuries Are The Most Common On The Jobsite. Read About How To Help Prevent Them In Occupational Health & Safety.
#ConstructionSafetyWeek #MilwaukeeTool #NothingButHeavyDuty
2005
Gsafety training companies/safety product service company7 May 2022 06:00:08Proper Head Protection could help prevent common yet serious injuries when workers slip, trip, or fall. Read More About the Importance and Types of Proper Head Protection in Occupational Health & Safety.
#ConstructionSafetyWeek #MilwaukeeTool #NothingButHeavyDuty
1974
Hsafety training companies/safety product service company16 September 2020 08:26:44#constructionsafetyweek Are you up to date with technology? One thing I like about DeWalt is they designed a line of power tools to provide a high level of dust containment control and low vibration without sacrificing performance. In return, keeping us safe. #safteyweek #preformprotect1816
IGeneral Contractor30 January 2022 12:18:21👷🚪🏡
#construcaocivil #construction #follow #likecommentshare #engineeringstudents #engineeringproblems #fibrereinforcement #viralpost #constructions #constructionequipment #civil_engineeringmemes #archidaily #civilengineeringdaily #indianarchitects #engineeringmarvel #civilengineeringlife #civilwork #civilengineerblog #reinforcedconcrete #civilengineeringdiscoveries #constructionsafety #constructionindustry #civilequipment #steelarchitecture #civilengineeringwork #indianconstruction #civilengineeringquestions #civilogy #civilpracticalknowledge #civilstructure
1810
Table 3. Influential users on construction health and safety on Instagram.
Table 3. Influential users on construction health and safety on Instagram.
UsersUsers’ Counts
Parallel Consultancy Ltd.282
Pike_Consulting_Group261
Global Training Canada197
Civil Techno Engineer193
Work Zone Coffee165
Canadian Safety Institute163
Inspiresafety1123
Ecoshield_Roofingnc105
Kenara.Multi Services98
Construction Safety Council97
Construcedllc94
Final Fit Safety91
Paintlynyc90
Core_Safety_Group90
Preferred_Safety_Products_Inc89
Fhconstruction89
Anvil_Builders85
Roland Construction_84
Ng Companies81
Vpg_Construction79
Table 4. Top 3 influential users on construction health and safety on Instagram.
Table 4. Top 3 influential users on construction health and safety on Instagram.
UsersFollowersThe Most Liked ContentLike Number Counts
Parallel Consultancy Ltd.351A party wall agreement is needed if you plan on carrying out any building work near or on a party wall. You must tell your neighbours, provide them with a Party Wall Notice, and come up with a Party Wall Agreement in writing.25
Pike_Consulting_Group115Many hazards can go unseen or looked over; simply double-checking can keep hazards from hurting people!⠀ #constructionsafety…7
Global Training Canada218Ouch. Knowing the hazards around you. #constructionsafety…7
Table 5. Source data.
Table 5. Source data.
+’ve Score-’ve ScoreNeutralNo SentNo DataOverallNet Sentiment =75.6%
Total Items Processed349340219491105000% Pos85.4%
Total Items Processed70%8%4%18%0%100%% Neu4.7%
% Neg9.8%
Average Number of Words per Item76838537 70Total #100%
No. Containing a URL or Email Address392452363 523
Percentage11%11%12%7% 10%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zeng, L.; Li, R.Y.M.; Yigitcanlar, T.; Zeng, H. Public Opinion Mining on Construction Health and Safety: Latent Dirichlet Allocation Approach. Buildings 2023, 13, 927. https://doi.org/10.3390/buildings13040927

AMA Style

Zeng L, Li RYM, Yigitcanlar T, Zeng H. Public Opinion Mining on Construction Health and Safety: Latent Dirichlet Allocation Approach. Buildings. 2023; 13(4):927. https://doi.org/10.3390/buildings13040927

Chicago/Turabian Style

Zeng, Liyun, Rita Yi Man Li, Tan Yigitcanlar, and Huiling Zeng. 2023. "Public Opinion Mining on Construction Health and Safety: Latent Dirichlet Allocation Approach" Buildings 13, no. 4: 927. https://doi.org/10.3390/buildings13040927

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop