This paper describes the development and new application of a text modeling process for identifying human factors topics, such as fatigue, workload, and distraction in aviation safety reports. Current approaches to identifying human factors topic representations in text data rely on manual review from subject matter experts. The implementation of a semi-supervised text modeling method overcomes the need for lengthy manual review through an initial extraction of pre-defined human factors topics, freeing time for focus on analyzing the information. This modeling approach allows analysts to use keywords to define topics of interest up front and influence the convergence of the model toward a result that reflects them, which provides an advantage over classic topic modeling approaches where domain knowledge is not integrated into the generation of derived topics. This paper includes a description of the modeling approach and rationale, data used, evaluation methods, challenges, and suggestions for future applications.
Understanding how underlying health conditions and social determinants of health affect the severity of COVID-19 is critical for community response planning. Literature reports that groups at higher risk from COVID-19 include those 65 and older, living in nursing homes and long-term care facilities, and with severe obesity, diabetes, chronic lung disease, or asthma. In addition, other studies has shown that the disease disproportionately affects individuals with lower socio-economic status. Our research seeks to validate these findings and observe the effects of health measures and social determinants of health on COVID-19 mortality at the county-level. In addition to COVID-19 research from hospital population samples, public health officials can leverage county-level factors for novel disease mitigation. We use the Johns Hopkins University COVID-19 reports of confirmed cases and deaths to measure disease mortality for each county in the United States. Then, we compare mortality to multiple county social determinants of health such as age, obesity, diabetes, and smoking in hypothesis testing. We fit multivariate linear models as well as non-linear models to predict mortality as a function of these county measures. The analysis shows that there is little evidence of a relationship between the county health measures of obesity, diabetes, or smoking and COVID-19 mortality as of the date of this publication. However, the analysis does reveal a positive relationship between the percent of a county population that is 65 or older and COVID-19 mortality. Other factors such as overcrowding, the percent uninsured, and the length of time since the virus has been detected in the county are also correlated with county COVID-19 mortality. Potential reasons for these findings, including data quality, are discussed. We also emphasize the advantage of collecting high quality, detailed health data at the county-level and explain how such data could be used to understand factors affecting the outcomes from novel diseases in real-time, as a disease is progressing.
We discuss our efforts in detection and tracking using Twitter data collected from January 2013 to the present and discuss various issues that arise in using Twitter data. We discuss various keyword methods, as well as methods for classifying a user as "sick". We discuss some of our successes and failures and provide some insight into the utility and limitations of Twitter. We discuss variations on the basic surveillance theme such as watching for a known disease, a known set of symptoms, and the more general problem of detecting an unusual number of sick individuals within a county.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.