
4 minute read
What is your salary level?
Furthermore, with a total number of tested samples put at 3,933,209 and confirmed cases at 249,154, Nigeria is one of the epic center of the virus in Africa accounting for 25,873 active cases, 220,195 discharged cases and 3,086 deaths as at 12th January, 2022 [4]. These figures and aforementioned factors trigger anxieties in the minds of the academia as they re-emphasizes the need for caution and proactive measures. Owing to the dreadful nature of the Omicron-variant, a fight or flight mental condition could pervade TEIs with far reaching implications on work disposition and since the fear of the virus could be more than the virus itself, the pandemic has loaded a huge psychological stress on people around the world, especially medical and educational front liners who have direct contacts with those tested positive due to the nature of their work [5]. Sense of impending danger amongst college staff, feeling of fear, panic and difficulty controlling worry will likewise be counterproductive especially for older staff with preexisting respiratory conditions [6]. Thus, there is dire need for a study that identifies causes of COVID-19 virus-scare in TEIs front liners not just to ascertain the true state of their mental health with respect to the dreadful Omicron variant, but to establish the links between data points and nosophobia, towards deducing informed conclusions for policy formulation by operators and regulators. This is the thrust of this study because when fear is prolonged or disproportionate, it could become injurious; hence the need to beam attention on the Nigerian workers with daily proximity to human traffic. Several methodologies have been adopted in literature for related works however, the adoption of machine learning as proposed in this study presents a more accurate predictive analytics of the fear stimuli, owing to its inherent efficient data studying capabilities [7]. Machine learning, through fuzzy logic, likewise helps in decision making processes and use cases including agricultural [8], medical, academics, etc. The results will help to identify most significant fear-stimuli and other germane associated topics in TEI communities would be unraveled through sentiment analysis and topic modelling methodologies. A more in-depth analysis of the academia-sourced dataset, through the instrumentation of machine learning and natural language processing, would give profound insight into opinions expressed by respondents. To the best of our knowledge, this study is the first deploying a three-throng computational approach to identify topical subjects in opinion texts towards unravelling fear-scare in academia-sourced data. Dataset acquired encapsulates feature attributes such as the demography of staff respondents, COVID-19 awareness level, personal health habits information, personal views about the pandemic, their environmental factors etc. The fear-stimuli predictive modelling is carried out on the Java-enabled Waikato Environment for knowledge Acquisition (WEKA, developed at the University of Waikato) and the natural language processing through Orange data mining toolkit (from the University of Ljubljana). Results are evaluated through machine learning performance metrics and discussed. The rest of the paper is structured such that session II discusses existing related works and literature review while session III unveils the methodology used for the proposed model. Session IV discusses the experimental result of the predictive analytics and session V concludes with recommendations.
II. LITERATURE REVIEW AND RELATED WORKS
Advertisement
Fear is commonplace in times of uncertainties like pandemic and it is an adaptive defense mechanism by animals central to survival and which involves several biological processes of grounding as a response to potentially threatening occurrences [9]. Identifying fear-provoking circumstances (stimulus) in work places in a pandemic will help tame anxieties and stress levels in healthy individuals and reduces the symptoms of those with pre-existing health disorders [9]. Similarly, it is opined that an important factor in understanding a population’s response to any threat whatsoever is the fear it elicits [10] since fear is a significant predictor of behavioral changes and health-securing behaviours [11]. Studies on earlier pandemics show that anxiety, or the lack of it, is a significant impulse of behavior [12] as people with slight anxiety about a viral epidemic are less likely to be involved in precautionary hygienic behaviors like periodic hand-washing, seldom observe physical or social distancing stipulations, and are less likely to take vaccinations if available [12]. Conversely, people with excessive anxiety are prone to socially disruptive behaviours including panic buying and frequent visits to hospitals, as minor symptoms are interpreted as signs of serious infections [13]. In [5], authors’ study focuses on assessing the psychological impacts of fear and anxiety amongst health front liners by conducting a single-center, cross-sectional survey through online questionnaires. Elements of fear, worry and melancholy were measured by the numeric rating scale (NRS) on fear, Hamilton Anxiety Scale (HAMA), and Hamilton Depression Scale (HAMD), accordingly. A total of 2299 appropriate contributors were consulted, which include 2042 medical staff and 257 administrative staff. The study observes that fear, worry and melancholy were significantly different between two groups and as compared to the non-clinical staff, front line medical staff with close contact with infected patients exhibit higher scores on fear scale, HAMA and HAMD, and they were 1.4 times more likely to fearful, twice more likely to be worrisome and depressed through melancholy. Similar to the work of [5], [14] undertook study on the impact of quarantine and physical distancing on mental health of participants as it seeks to find the nexus between the pandemic and the consequent containment measures whose result shows an association between higher levels of depression and anxiety symptoms amongst the surveyed population.