The study of data analytics in public health care application

Data analytics is the science of analyzing raw data to draw conclusions. In public health, data analytics refers to the use of statistical techniques and data mining tools to analyze large volumes of health-related information. It helps identify patterns, trends, and insights to inform decision-making, improve public health interventions, and monitor disease outbreaks for effective response and prevention. Unfortunately, most of the publicly available data is isolated, unstructured, insufficient, and invalidated. Moreover, employing data from various sources requires multiple processes before performing data analysis. The objective of this study is to demonstrate the processes of data analytics and how it can be applied to various data resources to answer public health issues that will help us better understand public health issues and guide us in managing public health policies. In this work, three sources of data have been considered. The first source was obtained from an automatic facemask-wearing detection system using artificial intelligence. The second one was the responses from an online survey. The last source was constituted from multiple public data sets that can be downloaded freely from the Internet. The first data source is artificial intelligence (AI) generated data. We explore the development process of AI-assisted face mask detection (AiMASK) during COVID 19 and how extensive data obtained from machine-based sources is utilized to extract valuable insights regarding mask-wearing patterns. This exemplifies the integration of AI and data science into public health, enabling the resolution of public concerns and informing the creation and adaptation of public health policies. The result of the study showed AiMASK has comparable effectiveness to human graders in identifying face mask usage. The increasing number of COVID-19 infections significantly influenced people's adherence to mask-wearing. A higher tendency towards not wearing masks was observed, particularly in the evening, during the holidays, and in the city centers. The second data source is in the form of primary data gathered through an online survey which acquired approximately 2,500 children. Following data labeling and cleaning, diverse data analysis techniques were applied to identify risk factors for computer vision syndrome (CVS) and determine the optimal duration of online activities to provide recommendations for adjusting the length of online classes for better ophthalmic health. During the lockdown, digital device usage among children increased, leading to more than 70% of them experiencing CVS. Factors such as the duration of digital device usage, online learning hours, ergonomics, and refractive errors should be adjusted to minimize the risk of developing CVS. The final data resource is an open data source of the national registry on road traffic mortalities (RTMs). We demonstrated data management of public data to analyze the mortality rates of road traffic injuries in children and adults over a decade. Additionally, we compared this data with the distribution of healthcare resources across 77 provinces in Thailand to identify any correlations or trends. We detected disparities in the distribution of road traffic injuries (Bateman & Sharpe) and the availability of hospital resources. Our study aims to contribute towards a more equitable reallocation of the resources, considering the varying numbers of traffic accidents in each province. In conclusion, diverse data sources require various tools and methodologies to generate novel public health data, thereby addressing the significant challenges faced by healthcare systems.