MSc. CIS Theses and Dissertations (2021)

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 5 of 6
  • Item
    A Web based Zakat collection and distribution system using K-Nearest Neighbors
    (Strathmore University, 2021) Samatar, Fatuma Abdullahi
    This research focusses on Zakat and how inefficiency in the process of Zakat collection and distribution impacts poverty. This research studies how the problem of Zakat management is handled in various parts of the world as well as takes a deep look into previous research and proposed solutions in order to come up with a system that attempts to improve efficiency and transparency of the process while building on previous research in the area. The researcher utilized the Agile methodology using a scrum approach to develop the system. The system included a front-facing rule-based calculator to improve the zakat collection process and a machine learning API, built using the K-Nearest Neighbors algorithm, to improve the efficiency of zakat distribution. As such, the model was built using the K- Nearest Neighbors algorithm as it outperformed the other common classification algorithm such as Decision Trees, Naïve Bayes and Support Vector Machine. This process leveraged existing libraries and tools in both the Python and the JavaScript ecosystem. This research concludes that inefficiency in the zakat process could be improved by systemizing the whole process and suggests the developed system as a starting point.
  • Item
    A Fuzzy expert based multi-criteria performance appraisal model for police officers
    (Strathmore University, 2021) Egessa, Bryan
    Performance appraisals play a key role in management of human capital and in particular in determining the effectiveness of an employee and to map out their developmental needs. Subjective appraisals are prevalent especially in the public sector, which has contributed to bias and inconsistent ratings and lack of transparency in determining promotions and reward. The appraised employees therefore hardly expect benefits from the exercise other than organisational compliance with regulations. Effective performance appraisal should be void of bias with measurable outputs that can guide various human capital management procedures. The purpose of this research was to establish the gap in appraisal practices in policing organisations and to present a solution that would enhance objectivity in the appraisal of police officers. Recommendations from previous research point to the need to have appraisal ratings arrived at from multiple sources that would include the public who are the primary consumers of policing services. Police satisfaction surveys and needs analysis reports have yielded recommendations that the public would provide valuable input in determining the effectiveness of a police officer. Other key players in determining appraisal ratings are the officer’s supervisor and peers with whom they have handled assignments. This research applied the prototyping methodology to design, develop and test a fuzzy expert based appraisal model with multiple appraisal data inputs and a single performance score output. To determine performance scores through a uniform procedure, a fuzzy controller was used to approximate the relationship between inputs and outputs via interpolation. The reason for applying fuzzy logic for the development of the appraisal model lies in the fact that vagueness is expected whenever human decisions are made. The crisp data inputs to the fuzzy controller were obtained from a mobile application developed to capture incident reports in real-time from the public, enable swift response by an officer and enable user rating upon completion of the assignment. Fuzzy logic is effective where multi-criteria decision-making is required such as an appraisal of employee performance. The web application provides an input interface for ratings by supervisors, peers and sub-ordinates as well as administrative tools. This model will enhance objectivity in the appraisal of police officers, which will also enable accurate identification of an individual’s development needs.
  • Item
    An Air quality prototype for monitoring greenhouse gas emissions
    (Strathmore University, 2021) Ngugi, Maureen Njeri
    In the world today, every human being wishes to live in a healthy, unpolluted and sustainable environment. This is because such a clean environment enables one to thrive and be productive in all aspects. Such environments are free from anything that may cause diseases and other physical injuries. Unfortunately, as years go by, our world has faced environmental degradation, global warming and high levels of pollution. This has not only affected wildlife and ecosystems in various parts of the world but it has also affected human health. This is evident by various respiratory diseases that have emerged such as pneumonia, bronchitis and many other diseases. This dissertation presents research work that focused on Green House Gas emissions which are a contributing factor to environmental degradation. It is important to monitor the amount of greenhouse gases in the atmosphere as it enables individuals, governments and environmental bodies to take action to tackle these emissions. This research used a prototyping methodology by developing an air quality monitoring system for greenhouse gases in the atmosphere. It incorporated an air quality monitoring prototype by integrating IoT with Wireless Sensor Networks. Collected data was then uploaded into a cloud platform using the Blynk API which relayed real-time information to a mobile device. The developed prototype achieved 95% accuracy. The developed systems can be used by individuals and environmental bodies to draw various strategies on how to lower Green House Gas emissions and adapt greener technologies that will be of great benefit to the environment as well as for a sustainable ecosystem.
  • Item
    A Mobile-based parent portal for public primary schools in Nairobi County, Kenya
    (Strathmore University, 2021) Ong’udi, Victor Okinyi
    Parental involvement in a child’s education and learner’s success are directly related. Because of this relationship, experts have developed different technologies to intervene and improve parental engagement efforts. For instance, the growing use of Internet and smartphones in Kenya has opened new and better technologies to improve parent involvement, especially in the public schools in Kenya. However, since many public primary schools rely on funding from the government and donations, it has been difficult for them to adopt technology in the learning environment because of a shortage of funds. Parents still have to make physical visits to schools to get updates on their child’s progress. Such challenges have impeded effective parental involvement in public primary schools. Hence, poor performance of learners and low public primary to secondary school transition compared to their private school’s counterparts. Therefore, affordable and easy to use products can be made to help the public primary schools achieve their academic goals and aid learning institutions to compete favourably with private schools. Thus, this study proposes a solution using a mobile based parent portal for public primary schools in Nairobi. The product will be developed based on the academic needs of the schools and reviews of the products that are already being used by learning institutions in other parts of the world. The parent portal will improve connection between schools and homes, enhance parental involvement, and ensure learner’s achievement.
  • Item
    A Deep learning-based system for de-identification of personal health information on mobile devices
    (Strathmore University, 2021) Musila, Daniel Mutiso
    Communication in healthcare has evolved from older technologies like pagers to present day smartphone devices. The change has been largely driven by the capability of smartphones to facilitate information exchange at greater speed and efficiency to manage the rising patient numbers, complexity of cases and the multiple disciplines in modern medicine. Instant messaging services like WhatsApp offer a channel which meets most of these needs. This communication often involves exchange of patient clinical data containing Protected Health Information (PHI). Various laws and policies have been enacted in various geographies and jurisdictions to safeguard the confidentiality of patients through strict management of PHI. During the normal course of care provision, healthcare professionals and organizations are expected to maintain full confidentiality and integrity of the data against unauthorized exposure. Whenever patient data needs to be shared with external parties for research use, informed consent must be obtained from the data subjects along with an oversight of their activities by a relevant review board. The widespread use of smartphones and popular instant messaging applications in modern healthcare however presents security and data protection challenges which need urgent addressing. De-identification of the data offers an avenue to address these concerns, allowing clinical data containing PHI to be shared among healthcare providers and/or researchers with minimized risks. Deep learning de-identification systems demonstrate superior performance over other approaches. They are generally deployed on high-end workstations in medical facilities and research centres, or on cloud-based infrastructure. However, on-premises deployments present infrastructural, connectivity and cost implications while cloud de-identification services may involve transmitting sensitive data across different jurisdictions therefore potentially breaching data residency regulations. On the other hand, smartphone use worldwide continues to see incredible growth with mobile processors becoming more powerful and versatile. Deep learning models can be deployed on Android-based smartphones to perform complex tasks such as de-identification of PHI. This is in line with the growing interest and research in edge computing, where computations are carried out as close to data sources as possible as an alternative to cloud computing. Concretely, this research proposes a mobile-based de-identification system, in which the deep learning model is optimized and embedded onto a smartphone application from which de-identification can be done. Specifically, Long Short-Term Memory (LSTM) artificial neural networks will be leveraged to develop a deep learning model which can then be ported onto the Android operating system to be embedded into a mobile de-identification application.