The 2024-25 Academic Year in Research: Part 2

The 2024-25 academic year saw a number of successes on the research front. Jumping to September of 2024, TSYS school computer scientist Yi Zhou provides a comprehensive overview of past developments and recent progress in the area of AI technology, which has developed rapidly in recent years, leading to widespread use in daily life. Zhou's study in Future Internet first summarizes the definition and characteristics of smart healthcare and then explores the opportunities that AI technology brings to the smart healthcare field from a macro perspective. The final two sections of the study categorize specific AI applications in smart healthcare into 10 domains and identify 10 key challenges these applications face. The 10 applications lie in the areas of disease prevention and prediction, diagnostic imaging, personalized treatment plans, virtual health assistance, remote patient monitoring, drug discovery and development, robotic surgery, electronic health records, behavioral health support, and clinical trial matching. The challenges include data integration and interoperability, large-scale data handling, real-time processing, model interpretability, continuous learning and adaptability, security of AI models, ethical AI design, integration with electronic health records, scalability, and underserved/remote areas with limited connectivity.

     Also contributing in September of 2024 is Turner College MIS professor Yoon Lee, whose study in Information Systems Frontiers acknowledges that imbalanced data sets are a growing problem in data mining and business analytics, and that, for example, the ability of machine learning algorithms to predict the minority class deteriorates in the presence of class imbalance. Lee's study proposes three methods based on a wrapper approach (i.e., a feature selection technique that finds the best subset of features for a specific machine learning model and domain) that combines the use of under-sampling with ensemble learning to improve the performance of standard data mining algorithms. Lee then tests the ensemble methods on 10 data sets collected from the UCI repository with an imbalance ratio of at least 70%. Lastly, the study also compares the performance of their ensemble method to two other traditional techniques for dealing with the imbalance problem and show significant improvement in both recall and the average of precision and recall.

     In October of 2024, Turner College marketing professor Sungwoo Jung published a study in the International Journal of Business and Management Studies investigating how logo changes affect sales and stock prices of restaurants. As the research explains, restaurant brands feature a variety of logos, such logotypes (i.e., logos that primarily focus on text) and combination marks (i.e., logos that primarily focuses on symbolism and text), which are the focus of the study. To explore logo changes within these categories of logos, Jung examines nine restaurant brands that have made logo changes: Papa John's, Sweetgreen, Kura Sushi, El Pollo Loco, Dunkin’, Noodles & Company, Burger King, Popeyes, and Tim Hortons. For companies with available stock price series, the statistical tests presented in the study indicate that simplistic logo changes did not significantly alter stock prices. However, some companies' sales revenues did significantly improve one to two years after a logo change.

     An October 2024 study in Review of Marketing Science by Turner College management professor Mark James investigates the influence of Anglicisms, which are words and phrases borrowed from English and inserted into a foreign language, on product appeal in Italian print advertising. James conducted two original tests involving potato chips, a convenience product that is frequently bought and inexpensive, and stereo speakers, a shopping product that is infrequently bought, expensive and requires greater purchase time/effort. Results of these tests showed no effect of Anglicisms on the relationships between perceived product differentiation, perceived price fairness, perceived product global-ness, or perceived product modernity and product appeal. However, Anglicisms were found to have consistently altered the relationship between perceived product risk and product appeal. Specifically, while Anglicisms decreased perceived product risk for potato chips, they increased perceived product risk for stereo speakers, suggesting that the impact of Anglicisms on perceived product risk can operate in an independent mechanism and be product dependent.

     Turner College management professor Johnny Ho also had a big October 2024 with his study in Diagnostics acknowledging that the recent advancements in artificial intelligence, especially convolution neural networks, have made the potential of automated medical image analysis nearly comparable to that of health professionals. In medical imaging, these networks can help medical practitioners diagnose diseases like COVID-19 or pneumonia by highlighting the suspicious regions in chest X-ray film. Many contemporary deep learning techniques only focus on COVID-19 classification tasks using X-ray images, while few attempt to make it explainable with a saliency map. Ho's study proposes a deep learning approach with image enhancement/cropping and data augmentation to enhance COVID-19 classification. The deep learning approach adds a multi-layer algorithm for improved visualization in X-ray images and to define and calculate a severity assessment index that measures infection severity. In testing, the trained model achieved an accuracy score of about 96.5% for the three-class X-ray classification task (i.e., COVID-19, pneumonia, and normal [healthy patients]), outperforming many existing techniques in the literature.

     An October 2024 study by Yoon Lee that appears in the International Journal of Production Economics asserts that when inventory includes perishables, the complexity of warehouse inventory management is compounded with additional requirements such as appropriate ambient storage conditions and placement of one type of perishable (e.g., bananas) far away from another type of perishable (e.g., strawberries). Lee's study utilizes analytical and simulation models to show that drone-based perishable inventory management is more efficient than manual perishable inventory management. The study adopts the Hotelling location model to build an analytical construct. The paper also creates a simulation analysis model to show the practical applicability of the model. Results from the analytical model and simulation analysis indicate that such warehouse automation is beneficial to both the warehouse operators and their customers.

     Turner College economist Frank Mixon closed out October 2024 with his study in Review of Behavioral Economics extending the constitutional economics literature by focusing on the recent and much discussed U.S. Supreme Court ruling to overturn Roe v. Wade and, therefore, the constitutional right to an abortion. The paper focuses on Justice Samuel Alito's now-famous 2022 draft opinion on Roe v. Wade, which, it argues, may have been leaked to Politico by a clerk to a conservative Supreme Court justice. According to the study, this possibility focuses on news reports indicating that Chief Justice John Roberts was working to convince Justice Brett Kavanaugh to join him and the liberal justices in upholding Roe v. Wade. As Mixon explains utilizing a game-theoretic vignette presented in extensive form, the leaking of Justice Alito’s draft opinion on Roe v. Wade can be seen as an unconditional strategic move referred to as "cutting off communication," which results in a player’s inability to receive messages from another player, thus committing the first player to a particular action. In this case, the leak served conservative justices by cutting off communication between Chief Justice Roberts and Justice Kavanaugh, thus thwarting the establishment of a case-specific coalition in this particular episode.