Conference Agenda

Stress experienced by a mother during pregnancy can profoundly influence the neurodevelopment of the fetus, potentially increasing susceptibility to autism spectrum disorders. We investigate these effects by extending a minimal mathematical model of the hypothalamic–pituitary–adrenal (HPA) axis to include mirror neuron activity in the premotor cortex. Our coupled system of differential equations captures the short-term (acute) versus long-term (chronic) impacts of maternal stress on embryonic development. We show that acute stress causes transient elevations in cortisol and mild alterations in mirror neuron activity, whereas chronic stress yields persistent physiological changes—elevated adrenal gland mass, higher baseline cortisol levels, and diminished mirror neuron activity that can remain even after the stressor is removed. Through phase-portrait and fixed-point analysis, we identify a key parameter, D, which governs the dynamics of mirror neuron activity. When D is small, three fixed points arise, including two stable equilibria—one corresponding to a “healthy” (euthymic) state and the other associated with reduced mirror neuron activity (an “autistic” state). Under large D, the system supports only the euthymic fixed point, indicative of a non-susceptible population. Finally, we demonstrate that while acute reductions in stress have negligible impact on the long-term dynamics, chronic reduction of stress can restore cortisol levels and mirror neuron activity to the euthymic range. These findings suggest that chronic maternal stress can be a critical determinant in the onset of autism-like neurodevelopmental trajectories for susceptible individuals and underscore the importance of sustained stress management interventions during pregnancy.

Integrating biological control with species conservation offers a sustainable solution for pest management while promoting biodiversity. This study examines a community-led conservation effort in Ayeyarwady Delta, Myanmar, where the invasive Golden Apple Snail (GAS) has severely damaged rice crops. The chemical controls and manual removal have been costly and environmentally harmful. In response, the Kyonekapyin-Tapseik Community Conservation Group (KTCG) in the delta area has focused on protecting the Asian Openbill stork (Anastomus oscitans), a natural predator of GAS, as part of a nature-based solution. We asked structured questionnaires to the farmers (n= 92) in eight villages. The results show a significant decline in GAS populations and a notable recovery in rice yields since 2018 when the stork protection has been started. Farmers observed that the storks’ foraging habits effectively reduced snail infestations. Before GAS infestations, average rice yields metric tonne per hectare were 0.73 in monsoon and 1.23 in summer, dropping significantly during infestations 0.15 and 0.85 in monsoon and summer, respectively, and restoring back to 0.55 and 1.21 in monsoon and in summer by protecting Asian Openbills which suppress GAS. The farmer attitudes survey (n=92) result shows that conserving biodiversity is important by strongly agree (n=89), agree (n=2), neutral (n=0) and disagree (n=1). None of farmers response for strongly disagree. This study highlights the dual benefits of conservation-driven pest control: improved agricultural productivity and species protection. By safeguarding the Asian Openbill stork, farmers not only reduced crop losses but also helped conserve a threatened species.

The rapid and continuous evolution of digital markets has significantly amplified the demand for intelligent, data-driven decision-making tools that not only enhance customer engagement but also improve operational efficiency and long-term business performance. In the highly dynamic landscape of digital commerce, businesses are increasingly faced with the challenge of managing large volumes of unstructured consumer data while making timely, insightful decisions. The complexity of customer interactions, particularly within e-commerce platforms, requires innovative solutions that can extract meaning, predict behavior, and support customer retention. In this context, the integration of Fourth Industrial Revolution (4IR) technologies such as artificial intelligence (AI), machine learning (ML), and intelligent automation has become indispensable for achieving competitiveness and driving strategic innovation.

This study presents a novel application of 4IR technologies, with a focus on machine learning and sentiment analysis, aimed at transforming digital business practices. Using Amazon e-commerce review data as a case study, we designed an intelligent framework that leverages advanced feature engineering, Synthetic Minority Oversampling Technique (SMOTE) for class balancing, and the AdaBoost ensemble learning algorithm. This integrated approach was developed to address key digital business objectives namely, improved sentiment classification and accurate prediction of customer churn.

We implemented and evaluated multiple classifiers including Random Forest, Logistic Regression, Extra Trees, and XGBoost, training them initially on baseline data and then enhancing them using SMOTE, AdaBoost, and their combined application. To comprehensively assess the models' performance, we employed several key metrics: accuracy, precision, recall, and Matthew’s Correlation Coefficient (MCC). Results demonstrate that AdaBoost significantly improves classification accuracy, with marked precision gains observed in Extra Trees and XGBoost. SMOTE was particularly effective in improving recall rates, especially when used with XGBoost. Notably, the combined use of SMOTE and AdaBoost resulted in enhanced MCC scores for both Random Forest and XGBoost models, indicating a more balanced and reliable predictive performance.

These findings underscore the transformative potential of integrating 4IR techniques into digital business systems. Our proposed framework proves valuable not only in enhancing the precision of sentiment analysis but also in enabling proactive management of customer churn two key challenges in digital business today. By providing actionable insights, the framework empowers organizations to make more informed, agile, and customer-centric decisions. This contributes to improved customer loyalty, reduced attrition, and more resilient digital business strategies.

Future research will explore the real-time deployment of the model in dynamic e-commerce environments and its extension to multilingual and multimodal datasets to ensure broader applicability and effectiveness across diverse global markets. This research ultimately contributes to the growing body of knowledge on digital transformation within the Industry 4.0 paradigm.

Medium-size and Large-bodied Mammals of Marshall Wetlands, a Proposed Protected Area in Liberia.

Abstract. Liberia comprises 43% of the upper Guinean tropical forest of West Africa, a global hotspot with high levels of biodiversity and endemism succumbing to accelerated rates of deforestation. Due in part to the country’s past 14 years of civil war, knowledge about much of Liberia’s remaining local biodiversity is limited. To help fill this gap, we surveyed medium-sized and large-bodied mammals in a proposed protected area, the Marshal Wetlands, in Margibi County, Liberia. We used 21 camera traps (Bushnell HD model 119739), installed at a minimum distance of 250 meters from one another, and at a height of 60 cm from the ground floor. Camera traps were set to take a sequence of 3 pictures every 2 minutes, using the highest quality setting both for the photos and sensor, to cover the entire trail. Trails included the three main habitats present in the area, primary and secondary forests, mangroves, and nearby human settlements. In an effort of 33,120 hours, and track surveys along trails, between July 2022 and May 2023, we confirmed the presence of 13 medium-sized and large mammal species. Of these, four species have some degree of threat in the IUCN Red List. Despite its proximity to the capital city of Monrovia, the Marshall Wetlands comprise significant biodiversity, highlighting its conservation value and reinforcing the value of improved protected status.

Key words. Biodiversity, camera-trapping, mangrove, species list, tropical forest, West Africa

Gweh A, Bailey A, Palmeirim AF, Poultolnor J, Toe H, Crowley W, Machalaba C, Desmond J, Desmond J,

Prist PR (2025) Medium-sized and large-bodied mammals of Marshall Wetlands, a proposed protected area

in Liberia. Check List 21 (1): 1–11. https://doi.org/10.15560/21.1.1