Kawan App is a mobile application built for IMCC USM. This app was never published due to difficulty in maintaining it by the stakeholders that do not have Xamarin developers. It is one of my early - learn as you go projects. This app is ready to use and some future work has been outlined for it.
A video demo is available on YouTube
A Kawan member is a student ambassador whose function is to help international students in Universiti Sains Malaysia (USM). Important questions related to and integrity problems in student ambassadors can be resolved through the integration of a data analytics component. This project is an extension of an existing Kawan web portal project with a focus on introducing new features comparable to social networking applications, volunteering applications, and on-demand service applications. The domain of the project is primarily focused around the Kawan initiative and involves, sign up, student-Kawan member matching, providing Kawan member services, logging activities, and providing feedback. The project consists of 3 major modules i.e. user authentication, profiles, and Kawan member services and includes a data analytics sub-module. The project was conducted in 2 iterations and included an experiment purposed to find the best supervised machine learning technique for the rank prediction task. The system developed within this project operates in a client-server architecture with two transmission modes. The system made use of the MVVM architecture pattern and it was concluded that Random Forest Classification would be the best supervised learning technique in terms of speed and accuracy for rank prediction. In total, 4 testing strategies were employed to ensure a smooth and proper operation of the system. Although the system achieved the desired expectations of the users, it could be improved by upgrading the Xamarin.Forms version to overcome a fatal bug and the system could be bettered in terms of compatibility. The goals and objectives of the project were met, but suggestions for future work are mainly centered in bettering the convenience of the system, employing more intelligent mechanics, improving the data analytics, and researching into transfer learning methods. The outcome of this project is a mobile application client that consumes a database and four servers. The mobile application makes available a data analytics dashboard with predictive analytics.
Keywords: Internationalisation, International students, Student ambassador, Kawan, Digitisation, Mobile application, Social Networking, Volunteering, On-demand services, Client-server architecture, MVVM, Data analytics, Random forest, Regression, Classification, Gaussian Naïve Bayes, Multinomial Naïve Bayes, Rank prediction
A video demo is available on YouTube