Publication Title 3D Imaging of Aquifer Structures in Agbor Using Very Low Frequency (VLF)Electromagnetic Data Download PDF
Publication Type journal
Publisher Journal of Computing, Science &Technology.Vol.1, Issue 1, 2024
Publication Authors Collins O. Molua
Year Published 2024-01-01
Abstract This study investigates aquifer structures in the Agbor region, employing Very Low Frequency (VLF) Electromagnetic data for 3D imaging. The research is prompted by the critical importance of aquifers in maintaining groundwater resources and the environmental repercussions of subsurface conductivity variations. The primary objective is to advance the understanding of aquifer structures through a comprehensive methodology. Focusing on Agbor, the study addresses the need for more detailed spatial data on subsurface conductivity variations, offering valuable insights for effective groundwater management and environmental assessments. The methodology entails designing and executing VLF electromagnetic surveys using a sophisticated VLF receiver along traverses. Rigorous data processing ensures high-quality measurements, including filtering, noise reduction, and signal enhancement. Inversion algorithms convert processed VLF data into resistivity-depth models, forming the basis for 3D representations. Geological data, such as borehole information and surface geology were integrated to refine the imaging process. Results are presented through scatter plots, line plots, and bar charts, showcasing electromagnetic signal variations, refined VLF data, and geological composition. Resistivity-depth models provide nuanced insights into subsurface resistivity variations, enhancing understanding aquifer systems. The findings bear practical implications for sustainable groundwater utilization and environmental studies in the Agbor region, addressing a significant knowledge gap. In conclusion, the materials and methods deployed encompass VLF electromagnetic surveys, rigorous data processing, and integration of geological data. The results offer detailed insights into aquifer structures, supporting recommendations for sustainable groundwater resource management, and environmental assessment in Agbor.
Publication Title A Framework of Deep Learning Based Terrorist Classification Model Download PDF
Publication Type journal
Publisher Journal of Computing, Science &Technology.Vol.1, Issue 1, 2024
Paper Link https://focjournal.unidel.edu.ng/
Publication Authors Adewale Olumide S.1 , Jimoh Ibraheem T. 2, Makinde Ibrahim A. 3 and Adeleye Samuel A. 4
Year Published 2024-01-01
Abstract Terrorism has claimed many lives and destroyed many homes throughout the world. All levels of government have made significant investments in security, but there has not been much progress in releasing the terrified public from the grip of terrorism. Several agencies have deployed kinetic techniques such as tactical apprehension and killing of terrorists but the menace keeps increasing, hence, the exploration of soft computing which use intelligence gathering, community engagement and collaborative approach to combat the monster jerking the world without using forceful approach against the terrorists. The aim of the work is to propose a framework that predict the terrorism activities and offer useful information to the security agencies to prevent the risks of terrorist attack. The work proposes the use the Bidirectional Encoder Representation from Transformer for the world embedding that will be feed into the Attention-Based Bidirectional Long Short-Term Memory to analyze the tweets from Twitter, a social network service, to predict the activities of the terrorist group and classify the terrorist group that is responsible for each attack and provide necessary information for both the security agency and members of the society. The framework will be implemented with Python, Natural Language Processing and Deep learning packages such as Keras, TensorFlow, sklearn, NumPy, Pandas, Natural Language Tool Kits while the evaluation metrics such as accuracy, precision, recall, specificity, and F1 score will be used.