Adetokunbo A. Adenowo1, Oreoluwa D. Jokosenumi2, Favour E. Onyebuchi3, Adam F. Zubair4

Introduction: Amid the rapid growth in digital media consumption on the internet, prioritizing efficiency, security, and user experience has become imperative in designing video hosting platforms. Aims: Hence, this paper presents the conceptualization, realization, and assessment of a cross-platform video hosting application named PlaySphere, which fulfills the criteria mentioned above. The motivation for this research came from the discovery of platform limitations (e.g., YouTube and TikTok). Such constraints in the platforms include disruptive advertisement, potential algorithmic influence, user data privacy, and restrictive and unfair revenue sharing monetization policy. Materials and Methods: The study utilizes a user-centered design (UCD) methodology. The methodology started with the initial stage of gathering data by the means of a Google Forms survey to thoroughly determine the user's requirements and preferences. The decision above led to the creation of a system architecture that was realised by: Firebase Authentication to help users register securely; Firestore to meet real-time data requirements; Firebase Storage to deal with media assets; and Paystack integration for dependable payment gateway and premium content access. Concerning data confidentiality and integrity, PlaySphere was set up to take advantage of Firebase's built-in AES 256 encryption for data at rest, along with Transport Layer Security (TLS) for data in motion. Functional testing of the complete system and user evaluations were carried out to determine the working, overall usability, and performance indicators. Results: The study results reveal that PlaySphere provides a smooth and enjoyable viewing experience, with low playback latency and good content management when contrasted with some traditional platforms. Furthermore, the assessment points to a high degree of user contentment, especially with the very simple look of the app, its stability, and quick responsiveness.

Taofik Ajagbe1, Adam Zubair2, Oluwatoyin Enikuomehin3, Mukaila Rahman4

Introduction: There is a belief in Islam that performing ablution before each prayer usually enhances the brightness of faces. In line with this assertion, there has not been proper scientific proof in determining facial conditions of Muslims in respect of the ablution effect to the brightness of their faces. Aims: To examine, investigate, and determine the strengths and weaknesses of the available machine learning algorithms for facial image analysis with respect to the effects of ablution on the human visage. Materials and Methods: The Google Scholar repository yielded 265 publications, all of which were thoroughly reviewed. After our predetermined inclusion and exclusion criteria were applied, only 48 papers were chosen for analysis in this study. Results: Results show that LBP, PCA, CNN, and SVM are the most frequently used algorithms for facial image categorization systems, and the accuracy of SVM ranged from 94.5% to 96.83%, and CNN returned an accuracy of 98.3%. Results also show there is a correlation between the size of input data and the accuracy of machine learning algorithms. Conclusion: This suggests that the Support Vector Machine (SVM) algorithm returned a higher accuracy of 96.83% in classification, especially when used with fewer datasets, and CNN worked robustly on larger datasets with an accuracy of 98.3%.