A new study published in Frontiers in Built Environment by John Ofori Datsa and William H. Collinge of the School of Civil Engineering at the University of Leeds set out to test whether automated sentiment analysis of Twitter discourse could reliably detect emotional patterns linked to occupational stress in the construction industry. The work is explicitly framed as a proof of concept rather than a snapshot of the sector today. The researchers mined a historical dataset spanning May 2008 to March 2016, a deliberate choice that allowed them to establish the technical feasibility of the approach without claiming contemporary generalizability. Their central finding is that social media platforms do contain measurable emotional signals about work experiences, and that these signals can be extracted and classified with reasonable accuracy using established natural language processing techniques.

The methodology began with careful data collection. Using Twitter’s X API V2 accessed through the Tweepy Python library, the team deployed six sequential search queries built around keywords and hashtags such as “construction,” “#constructionindustry,” “construction worker,” “construction site,” and “construction management.” Geographical filtering for UK-origin tweets was implemented through two complementary approaches: explicit GPS-derived coordinates falling within UK boundaries, and inferred location based on user profile fields matched against a predefined list of UK cities, postal code prefixes, and regional descriptors. Validation against tweets with explicit geocoding showed 94.2 percent agreement between the two methods, suggesting that profile-based inference is reasonably reliable. Because many users disable location services, relying exclusively on explicit geolocation would have excluded a substantial portion of UK discourse.

From an initial retrieval of 3,847 tweets, the researchers applied progressive filtering to arrive at a final analytical dataset of 2,322 tweets. Retweets were excluded to capture primary authorial intent, tweets shorter than 20 characters were removed as truncated content, non-English posts were dropped, and tweets discussing suicide or self-harm by third parties were excluded on ethical grounds to avoid amplifying harmful narratives. Preprocessing removed URLs, special characters, and stop words, generated tokens using the Natural Language Toolkit, and excluded tweets containing 50 percent or more non-dictionary tokens as likely noise or spam. The final corpus contained 42,344 tokens after stop-word removal, with 8,456 unique tokens and an average of 18.2 tokens per tweet. Notably, the researchers deliberately avoided lemmatization and stemming, preserving words in their original grammatical forms to retain semantic nuance for more accurate emotion detection.

Emotion labels were assigned using NRCLex, a lexicon-based tool derived from crowdsourced annotations by the National Research Council of Canada that maps 14,182 English words to ten emotion categories: anger, anticipation, disgust, fear, joy, negative, positive, sadness, surprise, and trust. Each tweet could carry multiple labels, with a mean of 2.3 emotions per tweet. To validate the automated labels, two independent human raters blind to classifier results assigned emotion categories to a random sample of 200 tweets. Inter-rater agreement between the humans was substantial at Cohen’s kappa of 0.82, and agreement between human ratings and NRCLex labels was 0.79, comfortably above the 0.70 threshold for acceptable agreement. This validation step is critical, as it establishes that the lexicon-based emotional classifications reasonably reflect human interpretation of the tweets despite relying on automated word matching rather than context-sensitive reading.

The machine learning phase was ambitious in scale. The team trained ten classifiers spanning diverse algorithmic families, including logistic regression, support vector classifier, random forest, decision tree, gradient boosting, XGBoost, CatBoost, AdaBoost, multilayer perceptron, and naive Bayes. Features were generated at three n-gram levels and weighted using term frequency-inverse document frequency vectorization. The dataset was partitioned using stratified random sampling into 80 percent training, 10 percent validation, and 10 percent test sets, with stratification preserving emotion class distribution across partitions. In total, 300 model instances were trained across all combinations of ten classifiers, ten emotion categories, and three feature types. Hyperparameters were deliberately left at published scikit-learn defaults rather than tuned, enabling reproducible baseline performance.

The results revealed a clear pattern: uni-gram features, meaning individual words, consistently outperformed bi-gram and tri-gram feature sets across all classifiers. This suggests that individual word choices carry sufficient discriminatory power for emotion detection in occupational discourse, without needing longer contextual phrases. For anticipation detection, logistic regression and CatBoost achieved performance comparable to the multilayer perceptron neural network, with AUC values of 0.93 versus 0.90. For fear detection, the multilayer perceptron achieved the superior AUC of 0.90. In the anger detection task, CatBoost, the multilayer perceptron, and the support vector classifier produced similar performance at approximately 0.76. Overall classification performance ranged from 0.76 to 0.93 AUC, with macro-averaged accuracies between 62 and 72 percent, substantially exceeding the naive baseline of 30.7 percent that would result from always predicting the most frequent emotion, trust.

Beyond classification, topic modelling using Latent Dirichlet Allocation identified ten substantive themes in construction-related discourse, with coherence scores peaking at ten topics. The three dominant themes, project scheduling and deadlines, employment opportunities and career advancement, and workplace safety and risk management, together accounted for 49.2 percent of the corpus. Employment security concerns were found to be embedded throughout technical discussions, suggesting that anxieties about job stability permeate even conversations ostensibly about equipment, scheduling, or management. By contrast, proactive discussions of occupational stress and organizational innovation received minimal attention, at less than 3.2 percent combined. This asymmetry is telling: workers discuss the operational pressures of their jobs far more readily than they discuss the psychological toll those pressures take.

The implications of this work extend beyond the construction sector. Traditional studies on worker wellbeing often rely on surveys, interviews, or organizational reports, which can be limited by response bias, restricted sample sizes, and the inherent delay between data collection and analysis. Social media analysis offers a complementary stream of naturally occurring text data that could enable more immediate, and potentially real-time, monitoring of workforce sentiment. By identifying patterns of negative sentiment, frustration, or stress within online discourse, organizations, policymakers, and industry stakeholders may be able to detect emerging workplace challenges earlier and develop proactive intervention strategies aimed at improving worker wellbeing and organizational productivity. The finding that default-parameter traditional classifiers match neural network approaches also has practical significance, as it suggests that meaningful emotion detection can be implemented without extensive model optimization or sophisticated architectures.

Important limitations temper these findings. Social media users may not fully represent the broader construction workforce, as participation in online discussions varies across demographic groups and job roles. Sentiment interpretation can be influenced by linguistic context, sarcasm, or ambiguous phrasing, which may affect classification accuracy. Most significantly, the historical timeframe of the dataset predates major industry transformations, including pandemic-era workforce disruptions and accelerated technology adoption. The researchers recommend contemporary replication using 2020 to 2026 data to assess whether emotional patterns have shifted with industry evolution and post-pandemic work adaptations, alongside expansion to multiple platforms such as LinkedIn and industry forums, longitudinal tracking of emotion signals, and integration with qualitative methods to contextualize the quantitative findings. As a methodological proof of concept, the study establishes that the emotional pulse of an industry can, in principle, be taken from the digital traces its workers leave behind.

Subject of Research: Social media sentiment analysis as a tool for monitoring mental wellbeing in the UK construction industry

Article Title: Investigating mental wellbeing in the construction industry using social media sentiment analysis: a UK case study

Article References: Datsa, J. O., & Collinge, W. H. (2026). Investigating mental wellbeing in the construction industry using social media sentiment analysis: a UK case study. Frontiers in Built Environment, 12, Article 1981301. https://doi.org/10.3389/fbuil.2026.1981301

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1981301

Keywords: mental health, construction industry, sentiment analysis, machine learning, social media, occupational wellbeing, UK, Investigating, mental, wellbeing, construction, industry

The study rests on an integrated theoretical framework combining three established models. The Job Demands–Resources model explains how excessive workloads, environmental hazards, and psychological strain erode well-being when they are not balanced by adequate resources such as protective equipment, supervision, and social support. Reason’s Accident Causation Theory contributes the distinction between latent failures, which are systemic weaknesses embedded in organizational processes, and active failures, the unsafe acts committed by frontline workers. Human Capital Theory completes the picture by framing education, training, and skill acquisition as determinants of productivity and safety. Together, these lenses allowed the researchers to trace how structural conditions, managerial decisions, and worker capacities interacted to produce harm during the crisis.

On the physical health front, the interviews documented a striking rise in respiratory illnesses, including asthma, shortness of breath, and chest congestion, which participants linked to the absence of personal protective equipment, overcrowded sites, viral exposure, and environmental pollution. Workers with pre-existing conditions were hit hardest. One asthmatic worker described having to continue living in a tent on site to support his family, which worsened his condition. Skin diseases such as itching, eczema, and infections also surged, attributed to frequent contact with unclean tools and limited access to safe water. Notably, the study captured a sharp divergence in accountability: building owners held contractors responsible for worker protection, while contractors framed the illnesses as natural phenomena or the result of workers’ own poor hygiene practices, and a government official acknowledged that Bangladesh lacks a comprehensive occupational health framework tailored to construction, with no standardized injury reporting or emergency response protocols.

Mental health emerged as a second, less visible casualty. Fear of infection combined with income uncertainty imposed by movement restrictions created profound psychological strain. The researchers found that socioeconomic position shaped behavioral responses: some workers from relatively well-off families simply stopped working to protect themselves, while those without savings faced an agonizing choice between infection risk and family survival. One worker recounted being penniless during the pandemic and continuing to work despite fear and uncertainty because stopping meant no income for his family. This psychosocial tension, the authors argue, represents a persistent dilemma for informal laborers whose livelihoods depend on daily physical presence at hazardous sites.

Environmental conditions at the worksites compounded these risks. Participants described unsanitary sites lacking adequate water and soap, makeshift tents that were dirty and overcrowded, and waste scattered across working areas. Weak site organization and poor management amplified the hazards. Yet even here, narratives diverged: one contractor insisted that soap and water had been provided at every site under his contract, blaming workers’ lackadaisical attitudes for non-compliance, while workers pointed to the absence of basic facilities. The researchers interpret this gap between managerial claims and worker experience as evidence of both weak enforcement and uneven adherence, mediated by limited education and awareness.

Workforce management failures formed the study’s second major theme. Contractors, who often serve as the sole authority on Bangladeshi construction sites, were frequently absent during the pandemic. Their absence produced disorganization, communication gaps, and neglected site maintenance, with debris left scattered and workers receiving no clear guidelines or schedules. In the power vacuum, older workers, those with higher education, and relatives of contractors assumed informal supervisory roles, pressuring colleagues to complete tasks quickly to demonstrate performance to the contractors rather than to protect health and safety. Compliance with protective measures proved almost entirely externally driven: workers reported wearing masks and gloves only when the contractor or site owner visited, focusing otherwise on finishing quickly to secure their daily wage.

The pandemic also fractured schedules and supply chains. Lockdowns, transportation restrictions, and shop closures disrupted the delivery of construction materials, while contractors’ absences left workers uncertain about their next tasks and procedures. Irregular working hours, sometimes beginning before 9 a.m. or extending past 5 p.m. without consultation, spilled over into workers’ family lives and mental health. Building owners reported difficulties paying wages and meeting financial obligations to suppliers and subcontractors as cash flow constraints intensified. The study links these operational breakdowns directly to increased occupational risk, arguing that role ambiguity and disorganized activity translated managerial inefficiency into physical danger.

Perhaps the most culturally distinctive finding concerns religious belief. Many workers interpreted COVID-19 as a curse from Allah, a consequence of human sins that would affect only the guilty, and believed that ultimate protection came from God rather than human intervention. While such beliefs provided psychological comfort, they often reduced adherence to preventive measures. One building owner described advising a worker to refit a mask that had slipped beneath his lip; the worker refused, stating that infection happens according to Allah’s will. The authors emphasize that these fatalistic interpretations, particularly common among workers with limited literacy, underscore the need for health education that is culturally sensitive rather than dismissive of religious context.

Skill development, the third theme, revealed a structural failure long predating the pandemic. Formal training programs and vocational certification schemes for construction workers in Bangladesh remain limited, and most workers enter the trade without task-specific competencies in plastering, bricklaying, handling sharp tools, or working at heights. Skills are instead acquired through informal mechanisms: learning from mistakes, observation, peer mentoring, and occasional safety meetings held by contractors. One worker described progressing from material bearer to plasterer over nine years of gradual development, while noting that many newcomers lose their jobs before achieving proficiency. The pandemic severed these fragile learning pathways, as lockdowns and social distancing ended safety discussions and reduced contact with experienced workers, leaving novices confused, error-prone, and unsafe.

The study’s integrative contribution lies in showing how these three domains reinforce one another. Limited skills made workers unable to understand or apply pandemic protocols, such as physical distancing or sanitizing shared tools, while managerial absence removed the guidance that might have compensated. The result was slower task completion, repeated errors, interrupted workflows, and reduced productivity, alongside heightened infection and injury risk. The authors propose a multi-level response: strict enforcement of the Bangladesh National Building Code, institutionalized safety audits, mandatory contingency planning with risk financing and insurance, continuous site-level supervision with quality protective equipment and digital communication tools, structured competency-based vocational training, and community-based mental health services and awareness campaigns delivered by NGOs and labor organizations. They acknowledge limitations, including the regional focus on Sylhet, purposive sampling, and self-reported data, and call for mixed-methods and post-pandemic research. But their central conclusion stands: the pandemic exposed how structural, managerial, educational, and cultural factors interact to produce occupational vulnerability, and only coordinated, evidence-informed action across all of them can build a resilient construction workforce in Bangladesh.

Subject of Research: Impacts of the COVID-19 pandemic on occupational health, workforce management, and skill development among construction workers in Bangladesh

Article Title: Impacts of COVID-19 on occupational health, workforce management, and skill development in the construction sector in Bangladesh

Article References: Ahmad, I., & Hossain, M. I. (2026). Impacts of COVID-19 on occupational health, workforce management, and skill development in the construction sector in Bangladesh. Discover Social Science and Health, 6(1), Article 94. https://doi.org/10.1007/s44155-026-00426-2

Image Credits: AI Generated

DOI: 10.1007/s44155-026-00426-2

Keywords: COVID-19, construction industry, occupational health, Bangladesh, workforce management, skill development, informal workers, mental health, workplace safety, qualitative research, Sylhet, labor informality