Masters Candidates

Research Supervisor: Dr. Evans Miriti

Project Title: A Veterinary Notification Service for Livestock Farmers(A case study of Bumula Sub-county, Bungoma County)

Abstract: The problem of timely access to Veterinary Doctors by Livestock farmers has paused great deal on livestock mobile health (mHealth) in Bumula Sub County, Bun Goma County. This problem is traceable at the initial stage of locating and communicating to veterinary doctor for the veterinary service. The project’s aim was to investigate application of modern Information Technology (ICT) in addressing accessibility to veterinary doctors by locating the nearest veterinary doctor and request for livestock treatment. The specific objectives that project sought to address include; Determining communication needs for the livestock farmers, propose & design Veterinary Notification Service prototype in which livestock farmers would locate and request the nearest Veterinary Doctor as soon as livestock fall sick. In this project, a case study of Bumula Sub County in Bun Goma County was used to study livestock farmers’ challenges in reaching out to Veterinary Doctor. Tests and evaluation findings indicated that application of integrated services of simple message service (SMS), Web application and Mobile application have positive outcome in veterinary notification process necessitated by mapping the nearest Veterinary Doctors to the livestock farmers’ treatment requests. The introduction of veterinary notification service for livestock farmers is a solution that would greatly improve in animal husbandry healthcare, treatment and real time consultancies in Bumula Sub County. Further study should also explore methods that can be used to facilitate other services on the prototype such as information and videolized basic livestock care.

Title: Assessment of Public Value of ICT Investment in County Governance: A case of Homa Bay County.

Abstract:

Public value is used when evaluating the performance of e-government. Furthermore, the performance of e-government is assessed from the citizen’s perspective. Transparency, participation, and collaboration in county governments’ processes should be viewed as means toward desirable ends, rather than administrative ends in themselves, as they appear to be currently treated. The creation of public value is the goal of public organizations; through public value, public organizations meet public goals with respect to substantive benefits as well as the intrinsic value of better government.Data was collected through questionnaires from residents, leaders and staff of Homa Bay County.Considering the nature of e-government in Kenya, four dimensions of public value creation through e-government were considered. These dimensions outline the four different ways of creating public value including the delivery of public services, the achievement of desirable outcomes, the development of trust, and the effectiveness of public organisations. The study recommends that a sensitization be done among the county residents of Homa bay on the existing electronic government channels they can use to access County services to allow more uptake of the ICT technologies for their own good. There is need to increase the services being provided through ICT e-government platforms to quicken up take of the ICT systems. This coupled with support on the usage and supporting systems such as infrastructure and power will help more services to be done through ICT platforms and more residents to use them. Going forward, the county needs to assess the public needs and how such needs can be handled through investments in ICT to increase the level of public value derived from the county’s ICT investments.

Project Title: A Big Data Analytics Architecture for Urban Transportation System: A Case for Nairobi City

Abstract: In the recent past, studies in the Intelligent Transportation Systems are on high gear to bring a solution to urban traffic systems caused by high rate of urbanization in most cities. Urbanization in Nairobi City has witnessed rapid growth over the last 10 years. Researchers have found that expansion of road networks does not solve the traffic situations. There is great loss of productivity for commuters who have to spend long hours in traffic. This research found that traffic information on updates, congestion and incidents is shared in various formats in various channels including radio, google maps and social media platforms. This information is big data and therefore the need to establish architecture for Big Data Analysis to synthesize and make use of data generated by road users freely through social media platforms. City travelers, find themselves in congested and jammed route due to lack of accurate and timely traffic insights of how the traffic is expected on certain routes to better plan and adjust accordingly. Policy Makers, Researchers, Systems Developers and Investors are also to benefit from the research findings.

Project Title: Smart Card Based Personal Health Records System for Interoperability

Research Supervisor: Prof Daniel Orwa

Abstract: Healthcare systems in both developed and the developing nations face numerous challenges such as lack of interoperability and limited personal medical records portability. Lack of conceivable personal health record’s interoperability has resulted in medical data silos. Patients on the other hand are seeking control over their personal health records. The research aimed at demonstrating the capability of a smart card based personal health records intervention using a prototype, guided by: - assessing the use of smart-cards in medical care; finding ways of giving patients more control over their personal health records; assessing the challenges healthcare workers experience in sharing personal medical data with patients and finally establishing the minimum data set vital for provisioning continuity of medical care.  The research concluded that: - majority of patients don’t have access to their individual medical health data; there is little use of smart health cards; medical tests are the most vital information needed for continuity of medical care, and that sharing of personal medical data is majorly manual physical files and word of mouth. Finally, the research concluded that the intervention suggested bridges the medical data silos and gives patients control over their personal health records. The research recommended that the government should actively participate in the implementation of eHealth projects and initiatives. Additionally, healthcare facilities should be more willing to participate in breaking data silos. Moreover, the use of smart health cards should be encouraged. Finally, patients should be encouraged to participate in their own health to help improve health outcomes.

Project Title: Development and Evaluation of an mHealth based Multi-Intervention Service for Adolescent Sexual and Reproductive Health Promotion

Research Supervisor: Dr. Samuel Ruhiu

Abstract: Young people are exposed to cultural content (from advertising to entertainment), that is highly sexualized, through the freely accessible World Wide Web. This has played a part in stirring up sexual curiosity and experimentation, leading to early sexual debut. The traditional cultural education and support system involving extended family and community has been severely eroded leaving the young people vulnerable to making inappropriate sexual choices, with accompanying negative Sexual and Reproductive Health (SRH) outcomes. mHealth provides a good global approach to health by the use of Mobile Phones; this links the young individuals to information concerning health and services. This study seeks to explore the feasibility and acceptability of a multi-intervention approach based on Mhealth to promote SRH among young people. A Persuasive System Model was used for developing a multi-interventional mHealth ASRH service. A prototype was developed that provided the following services: Adolescent Sexual and Reproductive Health (ASRH) education, Progress Track, Link to ASRH Service, Peer Support and Motivation. An online cross-sectional study was conducted with adolescents and young smartphone users aged between 14 to 24 years to evaluate the acceptability and feasibility of the intervention. A link to the mHealth prototype was sent to the participants who were required to interact with it for a period of 21 days. A total of 78 responses were received from the survey. Of the 85 participants recruited, 78 completed the survey. A majority of the participants noted that the intervention has a good design and that the information provided would help them make positive SRH choices. The results show the feasibility and acceptability of the intervention and its ability to influence adolescents' SRH choices.

Project  Title: Reducing Customer Churn in the Telecommunication Industry by use of Predictive Analytics

Research Supervisor: Dr. Evans Miriti

Abstract: Customer churn is a big problem in various businesses and especially so in the telecommunication industry. When a business loses its customers, it loses the revenue that was being generated from the customers and possibly revenue from potential customers who receive negative marketing from customers who churn. Managing customer churn in the Kenyan telecommunication industry has been largely ineffective due to the reactive approach where  churn is just a metric that is reported by the business after a certain period.The objective of this study was to show how we can use predictive analytics to proactively identify customers who are about to churn. By doing so businesses can take measures to prevent churn and therefore increase customer retention. This was done by identifying features that are most important in predicting churn, developing, implementing and testing churn prediction models. While there exist different approaches to solving the churn problem, machine learning was used to do the churn prediction based on various customer attributes such as age, usage, gender, etc. Since there exists multiple algorithms to do this kind of machine learning, this research implements four of them and does a comparison to see which one would be the most suited based on their performance. The final result shows which features can be used for churn prediction and their importance. The features were Registration Document, Age on Network, Subscriber age and Talk Time. The end result also showed how the different classification algorithms performed.

Project  Title: A Model for Processing Public Participation Feedback Using Topic Modeling (A Case for Public Task Forces in Kenya)

Research Supervisor: Dr. Lawrence Muchemi

Abstract: Governments acknowledge the need to involve citizens, through different platforms like public task forces, when public policies are drafted. This paper focuses on public task forces as one of the platforms for citizen engagement. The amount of feedback received is beyond the task forces’ processing ability which causes input from critical stakeholders being ignored completely leading to biased output. In this paper, we propose a model that task forces can use to process feedback received from stakeholders. Specifically, the model identifies topics contained in submissions by applying topic modeling using LDA algorithm. 22 submissions to an ICT procurement task force were used in developing the model. To validate the model, topics identified by the model were compared against those identified by a human expert. Results show that the model generated topics that are similar to topics identified by a human expert. This paper contributes to practice by enabling task forces to objectively identify topics covered by submissions which results into better acceptance of outputs and recommendations of task forces by the citizens. The model can be generalized to other establishments like constitutional commissions. The results of this study show that topic modeling can be applied in influencing the outcomes and quality of recommendations of task forces thus improving governance by ensuring that citizens’ views are adequately factored.

Project Title: Customer Segmentation on Mobile Money Users in Kenya.

Research Supervisor: Dr. Evans Miriti

Abstract: Customer segmentation enables organizations to partition a market into subsets that have common needs, interests and priorities. This helps businesses to come up with design and strategies that fulfils the customer needs.Using network and mobile money data, this study compared various clustering algorithms aiming at identifying the algorithm that creates the most solid customer profiles. Hierarchical clustering, KMeans and affinity propagation algorithms were used to segment customers and compared using internal validation measures. Our dataset comprised of various demographic and behavioural features obtained from a telecommunications company data warehouse. Co-relation between the features was tested enabling us to focus on age, network revenue, amounts transacted on mobile money, frequency of loan uptake, customer and organization transfers, goods and service payments and deposits and withdrawals as our features for modelling. The dataset was then fit into our algorithms. Agglomerative clustering generated seven clusters with a normalized mutual score of 0.5526 and adjusted rand score of 0.5436 and silhouette coefficient of 0.4523. KMeans generated 11 clusters with an NMI score of 0.5168 and adjusted rand score of 0.3315 and silhouette coefficient of 0.2369. Affinity propagation generated the largest number of clusters of 504, had a memory utilization of 91% and took the longest time to execute. This established AP as unsuitable for our dataset. Agglomerative clustering had the best performance in terms of the compactness and connectedness of clusters however clusters obtained from KMeans were more granular as compared to agglomerative clustering segments.

Project Title: Long Term Evolution Anomaly Detection and Root Cause Analysis for Data Throughput Optimization

Research Supervisor: Dr. Evans Miriti

Abstract: There is a growing demand for data which is driven by high number of smartphones, applications and traffic demand. Network operators have tried to provide enough capacity and meet the data speeds that the customer needs. This has led to introduction of new technology and expansion of the mobile networks making it complex to manage. Detecting anomalies that affect data throughput/speeds and investigating the root causes in mobile networks is challenging as mobile environments are increasingly complex, heterogeneous, and evolving. There is need to automate network management activities to improve network management processes and prevent revenue loss. Self-Organizing network is a standard introduced by third Generation Partnership Program (3GPP) to automate network management. However, the standard is still not fully developed. This project focused on implementing an anomaly detection and root cause analysis model that helps in the process of data throughput optimization in Long-term evolution (LTE) networks. The model usedDensity Based Spatial Clustering of Applications with Noise (DBSCAN) for anomaly detection, K-Nearest Neighbour (KNN) for root cause analysis and real network performance data from a Kenyan Operator. Proposed anomaly detection model achieved a silhouette coefficient of 0.451 showing a good separation of existing clusters in the dataset and was able to detect anomalies with both positive and negative impact on data throughput. The root cause analysis model achieved an accuracy of 94.59% and was able to identify the root cause of detected anomalies that had a negative impact on data throughput.

Project Title: A Market Basket Analysis Model to Address Visitor Cold Start Prediction Using Association Rules

Research Supervisor: Prof Elisha Opiyo

Abstract: E-commerce websites use recommender systems that recommend products with the highest user rating and those products that have similarities during customer interactions. These techniques give results from information retrieved during customer search and can be interpreted to match what the use is looking for. The recommenders simply respond to queries that customers invoke on the websites rather than customizing their responses to customer’s need. These recommender systems face multiple challenges like the cold start problem where a new user tries to purchase from the system or there is a new product that has no rating at all and therefore lack of enough information to recommend to the customer or the product, and an issue with overspecialisation. MBA approach with AR was proposed in this research to address the problem where the user has not interacted with the system before or there is insufficient information on the customer profile. The scope of the presented approach was to come up with a model that could accurately produce useful recommendations to the customer by identifying relationship patterns between products. The approach used Association Rules and predictive analysis of the products and intelligently identified similar sets of products that may interest a user so as to suggest those products that would satisfy a consumer more.

Project Title: A Multi-Tenancy Cloud Trust Model Using Quality of Service Monitoring: A Case of Infrastructure as a Service (IaaS)

Research Supervisor: Dr. Andrew Mwaura Kahonge

Abstract: Digitization and changes in technological trends have necessitated the need for enterprises to start or have plans to migrate their services to cloud computing environments. Thereby, benefiting from the many advantages that come with cloud computing. When a service level agreement is made between a cloud consumer and the cloud provider, the consumer usually is left with no choice but to trust the provider will deliver their bit. They have faith but lack a way of verifying or even doing QoS monitoring on their own. To solve this problem, we propose a multi-tenancy cloud trust model that uses QoS monitoring. It focuses on Infrastructure as a Service and, as seen in the results, assists cloud consumers evaluate cloud service providers well before they purchase services. This prevents them from leasing already congested clouds, or which do not meet their specifications. Cloud providers also stand to gain. A provider that honors their SLAs will tend to be more trusted than one who does not, leading to better reputation.

Project Title: Identity and Authentication Model for Bring Your Own Device in Organizations

Research Supervisor: Dr. Andrew Mwaura Kahonge

Abstract: The use of personal mobile devices for work related assignments and vice versa has become a common trend in business today. With escalated developments in technology, there has been a significant shift on how organizations and employees carry out business which has necessitated the adoption of new practices such as Bring Your Own Devices (BYOD). The research sought to review existing BYOD implementation models in device identity management and user authentication in organization and outline the challenges that face organizations in implementing BYOD strategies. It was found out that many organizations have implemented different models that address BYOD challenges from the perspective of the device, data and/or application. However, it was noted that illegitimate access to organizations’ systems via the use of legitimate devices and applications by unauthorized users remained a challenge that exposed organizations to unprecedented uncertainties. It was also established that the adoption of BYOD in organizations was low due to a number of reasons that were noted to centre on information security and privacy concerns. This research work developed and tested a prototype that combined device identification and user authentication in an augmented model as a proof of concept and a means of experimentation to address the security and privacy challenges that various implementations have had in the implementation and adoption of BYOD in organizations.

Project  Title: Digital Archiving and Preservation of Cancer Records: Case of KNH/UoN Department of Pathology.

Research Supervisor: Prof. Peter Waiganjo Wagacha 

Abstract: Cancer is considered to be 3rd leading killer and 2nd among non-communicable diseases in Kenya Mutinda J. (2019). Ferlay, et al (2013) noted there are about 37,000 new cases of cancer annually and an annual mortality rate of 28,000. This indicates cancer records rapidly accumulate over time and more resources are needed to collect and manage cancer-related data. KNH/UoN Department of Pathology have for a long time kept their records using a paper-based system. Objective of the study was to review current record keeping system including the process of preserving, storing and retrieving cancer records and the challenges faced, develop a model to guide the archival of locally available cancer paper-records and a web based prototype to preserve and avails this information to a wide range of stakeholders. The study used descriptive research design involving 22 participants. Random sampling technique was used to select respondents among the sampled population. The intervention prototype was developed using the RAD methodology. It was noted that cancer patient data were captured through standard forms/books (86% of the respondents) and stored in standard forms/books (71% of the respondents). 57% stated that identification of record and documenting was done based on LAB number. ICD 10 Coding system was not fully implemented as many forms had the old coding format. 77% of the participants stated that the inability to track patient records with ease as a challenge and lack of technology in records management as a potential security bleach and damage of records. Findings presented a strong case for this research study where, upon prototype development, there was concurrence among the respondents that the developed solution would be of significance to enhance cancer records management in the department and improve the healthcare service delivery process. The model can also be used to preserve other similar medical paper documents.

Project Title: Influence of Usability, Accessibility and Compliance on E-Government Development Index (EGDI): A Case Study of East Africa Countries.

Research Supervisor: Dr. Andrew Mwaura Kahonge

Abstract: Globalisation has necessitated the need for the promptness in information dissemination and access, regardless geolocation and time. Accordingly, countries have embraced Information Communication Technology (ICT) systems to widen accessibility to government services, reduce administrative cost, corruption, increase transparency, foster public participation, which have a positive impact on service delivery. Despite the level implementations and resource mobilization, these platforms still grapple with low-level adoption and usage as informed by the E-Government Development Index (E-GDI). The index assesses progress in online services delivery, open data and mobile services, and public involvement. The study sought to ascertain the influence of usability, accessibility and compliance on E-GDI using correlational research design.  Seventy-two government websites were selected from eighteen East Africa countries using a simple random sampling technique. Data was collected using Qualidator tool, Google mobile-friendly test tool, Nibbler tool, TAW Analysis Tool, Color Contrast Checker, and Readability test. The instruments deployed have a reliability of 0.98. Qualitative data were analyzed descriptively using inferential statistics and report generated by SPSS tool. Report indicated a significant correlation between usability, accessibility, and compliance on E-GDI. Based on the results, predictive equation was formulated for the dependent and independent variable.

Project Title: Use of Data Governance to Boost Corporate Performance in Kenya's SME Market

Research Supervisor: Christopher A. Moturi

Abstract: This study aimed to assess data governance practices and their effect on corporate performance within the SME sector in Kenya. This study used data governance decision areas based on Khatri and Brown, 2010 as a guiding model. Data was collected using questionnaires with participants derived from ten SMEs in Kenya. This research shows that data governance has an effect on corporate performance. It also shows that data quality, metadata and data lifecycle had a positive impact on data governance whereas data principles and data access did not. The study identified that customer satisfaction, operational efficiency, data quality and data security are some of the drivers for data governance with the SMEs. Consequently, barriers to data governance were identified as knowledge and experience, data ownership, culture, IT systems and unknown value of data governance. They should adopt initiatives that strengthen data governance such as clearly defining data roles and responsibilities, allocating resources to data projects and creating awareness through training. Data governance program for SMEs is necessary and proposed data governance framework can be used by practitioners to develop a data governance strategy and approach for managing data as an organizational asset. The finding of this study can be used to establish a data governance program for SMEs in Kenya. The suggested model can be tested and validated to determine its impact.

Project  Title: A Strategy to Reduce IT Risks in Technology-based Small and Medium-sized Enterprises

Research Supervisor: Prof. Elisha Opiyo

Abstract: SMEs in Kenya contribute greatly to the country’s growing economy. However, they face numerous challenges in sustaining their growth. A number of these SMEs are Technology-based, having technology and computing at the centre of their operations and performance. Hence, they face myriads of risks, throwing them into the drawing board to seek ways of managing such IT risks in their operations. The purpose of this paper is to identify a comprehensive risk management strategy, which will help in mitigation of potential IT risks in technology-based SMEs. This involved determining the key IT risks faced by technology-based SMEs and recommending the implementation of proper IT risk assessment and management methodologies.

Project Title: Digitisation Readiness Assessment in Public Organizations. A Case Study of Kenya National Examinations Council

Research Supervisor: Dr. Samuel Ruhiu

Abstract: An education system geared towards critical thinking, problem solving and lifelong learning is an important part of the innovation ecosystem. The examination process is an important indicator of learning outcomes. In Kenya, examinations are managed by Kenya National Examination Council (KNEC), who are required to provide accurate and timely examination information to education stakeholders, including candidates, schools, curriculum developers, and education policy makers. This is hampered by information held in hardcopy documents that poses a challenge to access, search, dissemination and analysis. A number of attempts to digitize existing documents have not been successful.This study purposed to develop a digitization readiness model for assessing the preparedness of KNEC and by extension other public organizations towards undertaking digitization. After reviewing theory on e-readiness and digitization of organizations, the study developed a digitization readiness assessment model (DRAM), which included organizational, IT governance, competency, technology and ICT security readiness indicators. The model was then validated through a survey at KNEC. Through a quantitative survey, the study sought to establish the preparedness of KNEC to carry out digitization. Purposive sampling targeting a population of 100 respondents was done. A questionnaire was the main data collection instrument, while data analysis was by use of frequencies, descriptive analysis and Principal Component Analysis.  The analysis established an aggregation of success factors along three components; most critical, critical, and less critical/supportive. The study reveals that the most critical indicators address governance of ICT projects, critical indicators addresses control measures of the same while less critical factors are supportive. The emergent digitization (preparedness) index for KNEC was found to be 2.88, on a scale of 1 to 4, where 2.5 is the minimum expected level of readiness. A critical look however at the individual indices that aggregate this score shows weakness in some of the factors associated with the “most critical” axis. These are competency readiness and Organizational Readiness. The study concludes that Digitization Readiness Assessment Model is useful to managers of public organizations, for decision-making and recommends sector wide approach towards digitization as a wayto optimize resources.

Project Title: Blockchain Readiness Assessment for Domestic Horticulture Traceability in Urban Kenya

Research Supervisor: Christopher A. Moturi

Abstract: There are numerous prospects that could be leveraged to transform horticultural supply chain traceability. The low usage, absence of regulations and policies, inadequate skills and awareness on Blockchain provides an opportunity for innovators, policy makers and researchers to explore measures of improving the readiness level and appropriate initiatives towards implementation of the technology. This study sought to investigate the existing traceability systems in Kenya’s domestic horticultural supply chain in order to assess the level of stakeholders’ readiness for the Blockchain technology. The Network Readiness Index was operationalized and used to extract constructs that were found to be relevant to this study. There is a potential for partnership between governments, producers and consumers to formulate the necessary regulations, policies and standards prior to implementation. The findings provide the actors with a basis to strategize on how to embrace Blockchain traceability systems in the horticultural supply chain.

Project Title: The Role of Mobile Technology in Dissemination of Information for Pensions Claims Processing

Research Supervisor: Prof. Robert O. Oboko

Abstract: Over the previous decade, Kenya has carried out main reforms in pension system in both private and public sector. One main reform among pension providers in Kenya has been to utilize mobile technology to improve information dissemination for pension claims processing. However, despite the use of mobile applications, short message services, USSD and voice calls, dissemination of information for pension claims processing still remains poor. This study deployed a descriptive study design. The target population was 561 pensioners and 8 heads of departments. The sample size was obtained by use of Krejcie and Morgan formula of determining sample size. The study employed systematic sampling method to choose 288 pensioners from the list of pensioners. The study utilized primary data as well as secondary data. Secondary data was acquired from the yearly reports. Primary data was acquired from key informant interviews and semi-structured questionnaires. Semi structured questionnaire produced qualitative and quantitative data, which was differently analyzed using diverse techniques. The key informant interview guide produced qualitative data. The study concludes that mobile application has a positive and significant association with dissemination of information for pension claims processing. In addition, the study revealed that text messages have a positive and significant influence on dissemination of information. Further, this study concludes that unstructured supplementary service data (USSD) has a positive and significant influence on dissemination of information. The study findings also revealed that voice calls have a positive and significant influence on dissemination of information for pension claims processing.

Project Title: Comparative Analysis of Distinctive Features of the Ransomware Tactics in Relation to other Malware.

Abstract: Ransomware have become a real threat to the use of technology. Unlike other forms of malware that could target systems by deleting or editing some files and creating backdoor for the attacker to access the system, ransomware have gone a notch higher by targeting humans. This is achieved when a ransomware encrypts data of the infected computer and a note demanding for a ransom to be paid is printed on the screen. Due to the advancement in technology, ransomware use advanced and secure encryption algorithm that is difficult to decrypt even when the computational power is not limited. In this work, we present some of the major behavioral characteristics that we found to be common with ransomware and not with other malware. Our results show that a careful analysis of suspicious network and file activities can help detect a ransomware attack. Further, careful analysis of ransomware behavior can help develop a system that can detect an impending ransomware attack and thereby eliminate it.

Project Title: Assessment of ICT Project Failure Factors in Government Agencies

Research Supervisor: Prof. Daniel Orwa

Abstract: Information Communication Technology (ICT) has become a critical differentiator in businesses enabling them to gain competitive edge over their competitors. This has led to emergence of e-governance which means that it is possible to provide efficient government administration and services through ICT. ICT Projects in the private sector do well compared to those in the public sector and worse in government agencies. However, in the past two decades, many governments around the world have become aware of ICT potential in enhancing their service and efficiency.  However, government ICT projects in developing countries are well known for running far behind schedule and failing to deliver the expected benefits. The purpose of this research was to identify the ICT project failure factors in government agencies by assessing the effect of project, people and technology related issues. The research used quantitative and qualitative methods as the research method. The research used questionnaires as the primary source of information coupled with informal interviews that were guided by the questionnaires. The research revealed that vision and strategy, project management, skills, solution design, software and hardware contributed significantly to the failure of an ICT project in government agencies. The research recommends presence of a clear vision and strategy in ICT projects, formulation of project management policies to govern project management teams across all government agencies, and creation and strict adherence to recruitment process that will ensure absorption of required skills in the project team and consequently decrease ICT projects failure across all government agencies.

Project  Title: Factors influencing adoption of electronic health management system in public hospitals in Kenya: A case of level iv hospitals in Nairobi County.

Research Supervisor: Prof. Robert O. Oboko

Abstract: Health Information management system is one of the most potent investments a health institution can effectively improve its ability in the provision of health services to its patients. However, the adoption of electronic health management system still remains low which is in contrast to what the ministry of health expected and only about 7 percent of the health care practitioners use them. This study therefore seeks to investigate on the factors influencing adoption of electronic health management system in public health facilities in Kenya. The study also seeks to assess the effect of the staff information and communication technology proficiency, electronic health management system functionalities, top administration support and resource availability on adoption of electronic health management system in public hospitals in Nairobi County. This study will use a descriptive research design. The target population of this study will be all the 235 staff working as medical doctors, nurses, pharmacists, clinical officers, laboratory technicians, accountants, administrators and registrars in level four hospitals in Nairobi County. These hospitals include Mbagathi District Hospital, Mutuini Sub -District Hospital, Pumwani Maternity Hospital and Lucy Kibaki District Hospital.

Project  Title: Using ICT to Improve Access to Educational Resources by Marginalized Communities: a Kenyan Experience

Research Supervisor: Christopher A. Moturi

Abstract: The study aims to develop an enhanced ICT-Based access model which will avail current and relevant educational resources, hence transform the lives of marginalized communities. Applying the PRIMER model, data was collected through questionnaires from 218 staff among the 30 technical and vocational colleges within the selected county in Kenya. The research found that the high cost of reading materials, low number of qualified trainers and lack of necessary ICT infrastructures to be the key challenges. The use of ICT based models, ICT policies and standards, community involvement were found to be the key access solutions. A model for improving access to educational resources, which can be generalized to other underprivileged societies in Kenya, is proposed. The study provides an alternative and practical solution to educational challenges within technical and vocational education and training settings for rural and marginalized communities. This will assist education stakeholders in their allocation of resources to TVETs in marginalized zones. The TVETs in Kenya plays a critical role, noting that only 18 per cent of students who graduated from form 4 in 2019 secured university placement. The study projects the critical need to improve access to educational resources for marginalized communities by address the inherent disparities in such settings which requires the adoption of suitable and beneficial technologies with a strong emphasis on the peculiarities of the community.

Project Title: An Assessment of Green ICT Implementation Benefits to Kenyan Organisations: a Possible Motivator for Adoption Of Green ICT

Research Supervisor: Prof. Agnes Wausi

Abstract: There has been an increase in the use of computers and computing devices in Kenya and generally in the world due to the reduced cost of ICT equipment and improved technological advancements. This increased use of ICT, despite bringing positive effects to an organization, does also pose some environmental and monetary negative effects on an organization and the planet. Studies have shown that ICT contributes to 2% of the world's greenhouse gases and in addition ICT resources have increased the overall consumption of electricity within organizations. Most computer hardware also contains hazardous material like mercury, cobalt, lead, etc. This study's objective was to establish if the expectation of green ICT benefits could be a motivator for adopting green ICT practices within an organization. The research also sought to establish some explicit benefits of green ICT enjoyed by organizations that practice green ICT. The methodology employed for this study was a purposive stratified research design. A study of organizations within the Nairobi metropolitan was conducted and results were analyzed from a population that constituted organizations from different sectors. Government, private, and not for profit organizations, with a majority of respondents being from the government. Due to challenges arising from the COVID 19 pandemic data collection was restricted to online questionnaires and secondary data from literature review. Study findings pointed to the availability of benefits of practicing green ICT within organizations and at the same time also shed some light on factors that do not persuade organizations to practice green ICT. The overall study finding greatly highlighted a strong correlation between organizations practicing green ICT and anticipated monetary benefits that result from the practice.

Project Abstract: The application of digital mobile forensic investigation has become an integral process both in private and public institutions. This has led to the development of various forensic tools in order to bridge the gap in the business processes. The development however has come at a cost; most of these tools are heterogenous and hence pose the challenge of interoperability. Nugget addressed this in the spectrum of digital forensic. The development however, was limited to disk, memory and network analysis. The exception of mobile forensic introduced an area of research, which after being addressed introduced a new gap, the application of mobile forensics into Nugget did not address the last two phases of mobile forensics which are analysis and reporting. To address this, we incorporated mobile forensic analysis and reporting into Nugget. The output of this research was a software tool that can be used by forensic investigators when performing mobile forensics right from extraction to reporting.

Project Title: Replication as a Way to Achieve Interoperability in HealthCare

Research Supervisor: Prof. Daniel Orwa

Abstract: Interoperability is one of the key information exchange platforms among systems that are running differently. Presently, Electronic Health Records (EHR) systems are developed differently by vendors for individual healthcare facilities and they therefore cannot communicate with other healthcare facilities in order to share the patients’ data. This makes it difficult for healthcare providers to share patient data and also to access medical history which is oftentimes used as evidence based in decision making at all levels of the system especially at the point of origin. This research sought to investigate some of the barriers that hospitals and other healthcare facilities face in the journey towards interoperability. To achieve this, the author looked into ways on how developed countries and other countries that have been able to share patient data among the different healthcare facilities do so and some of the technologies they use to achieve interoperability.  Findings indicated that currently, systems do not allow sharing of patient medical records instead patents undergo similar tests and activities in all the healthcare facilities they visit. The research developed a platform that uses replication through federated and query based data exchange model as a way of exchanging patients’ medical records. The platform allows healthcare providers to retrieve patients ‘records that are stored in different healthcare facilities.

Project Title: Forensic Analysis of Dropbox Data Remnants on Windows 10

Abstract: Cloud storage services are popular among businesses and individuals as they offer convenience in storage and sharing of files at an affordable price. However, cloud storage is subject to abuse by cybercriminals, and coupled with the difficulty in getting artefacts of evidential value from cloud storage providers, artefacts from client computer can provide potential evidence on which a case can be based. This research investigates artefacts left behind by Dropbox, a popular cloud storage application, on Windows 10. Through live and dead forensics, the study determines Dropbox artefacts on Windows 10 for various scenarios including installation, file upload, file deletion, and uninstallation. By identifying these remnants, this work contributes to a better understanding of the artefacts that are likely to remain for digital forensics investigators. Potential information sources identified during the research include the Dropbox client software installation files, synchronisation folder, browser, link files, prefetch files, registry, and network traffic. The artefacts identified in the study can assist in criminal investigation involving Dropbox as they provide useful information in recreating the scene of crime, tying a suspect to the crime, and creating a timeline of events.

Project Title: Telepsychiatry for the Treatment of Depressive Disorders in Kenya

Research Supervisor: Prof. Peter Waiganjo Wagacha

Abstract: Majority of the patients suffering from mental disorders do not have access to efficient and quality healthcare due to limited resources. With less than 100 psychiatrists in Kenya, 75-85% of the population does not receive the mental healthcare they need.  Scenarios where access to mental healthcare is limited or non-existent can aggravate mental issues further but this can be overcome with the use of telepsychiatry which has the potential to provide a platform for effective interaction between a client and healthcare provider. A review and analysis of existing models was utilized to develop a telepsychiatry system design model which was validated through the design, development, test and evaluation of a telepsychiatry prototype. For the evaluation, mental health experts used the system and later filled a post-hoc questionnaire. Research results established that limited access and financial restrains were found to be the major hindrances to accessing mental healthcare in Kenya and the prototype system provides a platform where a psychiatrist is able to remotely consult through chat and integrated video with the client, carry out standard assessment tests and take down personal notes. It also allows for easy storage and retrieval of client files particularly for monitoring purposes and the psychiatrists expressed willingness to take up the system if it is availed. In conclusion, the importance of such a platform in avoiding interruption of mental healthcare provision is timely especially in situations where people have restricted movement such as during the covid-19 pandemic.

Project Title: Evaluating the Impact of E-Government Services: A Case of Nairobi City County Electronic Payments Service

Abstract: Currently, most organizations and businesses are adopting technology as a means of offering quality service and transactions. The public service and more so government institutions are not left behind in embracing technology in order to offer services to the citizen. The Nairobi county government implemented an e-payment service (e-Jiji pay) to improve on its service delivery, where the users directly interact with the system. However, the users complained of system unavailability, tedious processes, and public awareness. This study adopted a descriptive research design, to obtain views from Citizens in regard to the effectiveness of NCCG portal guided by the independent variables identified in the research tool. The study was conducted on a sample size of 400 respondents. Questionnaires were administered to collect the data on citizens’ perspectives. A Cronbach’s alpha of 0.875 was obtained in order to establish the internal consistency of data. The results showed that the cost of service, service quality and user satisfaction has an impact on e-government, while, e-readiness was seen as not having a significant value on e-government services. Every sector thinking of deploying a successful e-service for the citizens must consider the cost of accessing the services, quality of e-services, and User satisfaction to those e-services. The county managers should, therefore, consider a better quality of e-services by enhancing; a user-friendly interface, information quality, 24/7 availability of services, response time, and have a variety of access options.

Project Title: Human Factors Affecting Favourable Cybersecurity Culture: A Case of Small and Medium-sized Enterprises (SMEs) Providing Enterprise wide Information Systems Solutions in Nairobi City County in Kenya.

Abstract: Recent news coverage in both print and electronic media clearly indicates that cyberattacks are increasingly on the rise. As compared to large enterprises, SMEs are highly vulnerable to cyberattacks for they lack adequate cybersecurity controls in place to cope up with evolving cyber threats. Reports from the industry have underlined human factors as the root cause of many cybersecurity incidents in organizations. This study, thus, purposed to examine the key human factors that impact on favourable cybersecurity culture in Kenyan SMEs that provides enterprise wide information Systems solutions and that are premised in Nairobi City County. To achieve this, a quantitative research inform of descriptive research design was conducted. Mail survey method was adopted to gather primary data through a structured questionnaire from SMEs selected from the official 2019 yellow pages Kenya online directory. The regression analysis results established that top management support and involvement together with reward and deterrence measures are positive and significant predictors of favourable cybersecurity culture and thus form important strategies for instilling favourable cybersecurity culture in SMEs considered. Other strategies that need to be developed by these SMEs are cybersecurity policy, cybersecurity change management, cybersecurity training and awareness programs, cybersecurity monitoring and audit for they were also found to have a positive effect on favourable cybersecurity culture. The study concludes by emphasizing the need for adequate and consistent top management support and involvement in cybersecurity issues.

Project Title: Supervisory Control and Data Acquisition (SCADA) System Live Memory Acquisition for the Modbus Protocol Forensics. A Case of the Petroleum Depots in Kenya

Abstract: Supervisory Control and Data Acquisition (SCADA) has been at the core of Operational Technology used in industries and process plants to monitor and control critical processes, especially in the energy sector. In petroleum sub-sector, it has been used in monitoring transportation, storage and loading of petroleum products. It is linked to instruments that collect and monitor parameters such as temperature, pressure and product densities. It gives commands to actuators by the use of the application programs installed on the programmable logic controllers. Earlier SCADA systems were isolated from the internet, hence protected from CYBER attacks. Recent trend in SCADA systems is integration to other business systems using Internet technologies such as Ethernet and TCP/IP. However, TCP/IP and web technologies which are predominantly used by IT systems have become increasingly vulnerable to cyberattacks that are experienced by IT systems such as malwares and other attacks.  It is important to conduct vulnerability assessment of SCADA systems with a view to thwart attacks that can exploit such vulnerabilities. Where the vulnerabilities have been exploited, forensic analysis is required so as to know what really happened. This research reviews SCADA systems configuration, vulnerabilities, and attacks scenarios, then presents a prototype SCADA system and forensic tool that can be used on SCADA. The tool reads into the PLC memory and Wireshark has been to capture network communication between the SCADA system and the PLC.

Project Title: Leveraging Robotic Process Automation as an Enabler of Organizational Digital Transformation

Research Supervisor: Prof. Elisha Toyne Opiyo

Abstract: Organisations are adopting newer technologies in a rapidly evolving digital environment to develop new or change present business processes to meet changing business and customer expectations. Robotic Process Automation is at the forefront of these disruptive technologies and has immense potential for transforming the delivery of processes. Organizations spend a significant amount of time managing processes that are highly repetitive, straightforward, rule-based and time-consuming and can be automated with RPA. However, often organizations which have embarked on their digital transformation journey, using RPA as an enabler, do not scale beyond concept proof. The research aimed at establishing a model for RPA adoption. Previous research on technology adoption frameworks, digital transformation frameworks and concepts on RPA supported the study. Influence of RPA performance expectancy, RPA effort expectancy, RPA facilitating condition, RPA social influence and RPA risks and threats on leveraging RPA were investigated with age, gender and experience as moderating factors. The outcome presented a framework for leveraging RPA.

Project Title: Effects of institutional pressure on BYOD information security compliance: A  case of two private and two public Kenyan universities

Research Supervisor: Dr. Samuel Ruhiu

Abstract: The rising acknowledgment of the importance of information security has created institutional pressures on universities to comply with Bring Your Own Device information security standards and policies for protecting their information. How such pressures influence Bring Your Own Device information security policy compliance in universities, however, is unclear. This research presents an empirical study to investigate the effect of institutional pressure on Bring Your Own Device information security policy compliance. With the use of multiple regression analysis for analyzing the data collected through an online survey, the study shows that coercive pressures, normative pressures, and mimetic pressures positively influence Bring Your Own Device information security policy compliance in universities. It reveals that institutional pressure achieves a higher percentage in Bring Your Own Device information security policy compliance if mediated by information security awareness. Theoretically this study contributes to information security research by better understanding how institutional pressures can be used for enhancing Bring Your Own Device information security policy compliance in universities. Practically this study informs information security policy makers of the major institutional drivers for Bring Your Own Device information security policy compliance.

Project Title: Convolutional Neural Networks to Enhance Stock Taking

Abstract: In the current competitive economic environment, there is a lot of focus in optimization of processes and providing high quality customer experience. This study explores the use of deep learning particularly convolutional neural network to enhance the retail store stock taking process. It provides a review of literature on different convolutional neural network architectures to identify the best fit for image object detection and count. It highlights some of the image analysis applications in various sectors such as counting fish, yield estimation and construction site management. YOLO is noted to be perform well based on the literature review and the study further implements and compares the performance of YOLO v2 and YOLO v3 in object detection and count. The implementation leverages on the pretrained weights on ImageNet and further training is done on open data image set of retails stores. Both YOLOv2 and YOLOv3 achieve mean average precision above 75%, however YOLOv3 is leading attaining a mean average precision of 81.86%.

Project Title: Early Detection and Prediction of Diabetes in Patients

Abstract: Diabetes Mellitus is a chronic non-communicable illness (NCD), that happens when the pancreas produces not enough insulin, or when it over produces the insulin that the body is not able efficiently use. There are more than 8,700 diabetes related deaths recorded in Kenya in 2015. It is estimated that diabetes causes 3% of deaths in Kenya. The prevalence may rise to 4.5% (approximately over 380 million people worldwide) by the year 2025. Most diabetes patients cannot access healthcare services; therefore, the conditions are not treated for long which leads to either deaths or permanent disability, either through blindness, kidney disease or amputation. So early diagnosis of diabetes is vital, in helping with on time treatment, thereby decreasing deaths caused by the disease and help medical practitioners in better decision making. The study’s main agenda was to identify the key factors to be used in predicting diabetes, design a prediction model using ML (machine learning) algorithms, and select the suitable model to give the best results. The result shows the Random forest algorithm had the highest accuracy of 75.2%, held best for the analysis of diabetic data.

Project Title: Use of Blockchain In the Informal Distributed Manufacturing Industry in Kenya

Research Supervisor: Prof. Robert O. Oboko

Abstract: In the last decade, research and development around distributed ledger technology (DLT) has grown exponentially. The financial services industry has been revolutionised by the explosion of cryptocurrencies like Bitcoin and Ethereum. Researchers have taken the principles used in these cryptocurrencies and are using them to develop other DLTs in various fields. This study explores how blockchain can be used to provide traceability, visibility and transparency in the Kenyan informal distributed manufacturing industry. SOKO, an aggregator of artisans spread all over Nairobi, was the case study used. Purposive and convenience based were the sampling methods used. SOKO supply chain employees and active artisans were the sample population. Interviews and observations were data collection methods used. Content analysis, a qualitative data analysis method, was used to capture emerging and predetermined themes. Google sheets and Dovetail were the tools used for this study. This paper finds that the use of the immutability and proof of origin features of blockchain greatly enhances traceability within a supply chain. It is imperative that the granular information collected should be intuitive and accessible to all parties to enhance visibility. Amplified traceability and visibility greatly improved transparency and accountability within the SOKO ecosystem. A pivotal recommendation for future research is usage of unit-based tagging technologies e.g barcodes, QR codes, or RFID. Combined use of such technologies and blockchain would achieve the highest level of traceability, especially when working with diverse producers who produce similar

products.

Project  Title: Prediction of Stock Prices Using Predictive Data Analytics: A Case of Nairobi  Securities Exchange

Abstract: The field of stock trading worldwide is not even, has never been even and will not be even in the years to come. The money bags, the big boys continue to benefit and make massive profits from the stock market and the small fish continue to lose millions and billions of money in the market.

Thanks to technology and the fourth industrial revolution, thanks to big data, predictive analytics and more so Holt Winters Exponential Smoothening, it is now very possible to predict the volatility at the stock market and to make significant gains like the big boys. With the predictive analytics, the massive stocks data from the stock market becomes a goldmine that generates leads for short term investors who are able to be warned in advance of the possible capital erosion due to their bad investment stakes at the market and be warned of the possibility of capital gains before investing in the stock market. The data used in this project is a ten years’ data and hence the predictions is 90% accurate. While it is impossible to predict with 100% accuracy what will happen in the future since stock volatility is affected by many other economic, political, environmental factors, it is very much possible to use historical data from the Nairobi Securities Exchange, do stocks prediction and predict with accuracy the future stocks movement if all other factors that affect stock volatility are held constant. The  constant, is one of the major assumptions in this project.

Unlike in the olden times where stock traders and stock market analysts would rely on fundamentals of a given company to decide whether to buy or sell, or rely on technical analysis by following graphs and trends, predicting the stock market is getting easier using predictive analytics. In the past, predicting the stock market volatility has been done by several technologies including Artificial Neural Networks, Decision Trees, Logic Regression, Ensemble Models, Bayesian Models among many other statistical models. This project looks specifically at Holt Winters Exponential Smoothing, a branch of supervised machine learning model.

A Web service based Integrated fitness and Diet Mobile Application For  Health Promotion

A sedentary lifestyle, unhealthy food and work related stress are part and parcel of the lives of many people thus their wellbeing and consequently their quality of life is adversely affected. Studies show that physical inactivity and improper diet contribute to over 3 million deaths annually. Efforts have been made in different areas to address the problem whereby different technologies such as web services in distributed systems, cloud technologies, internet, Artificial intelligence, mobile technologies have been used in providing solutions as interventions to lifestyle diseases.The emergence and pervasiveness of smartphones and mobile devices has led to an increase in health and fitness applications which possess an assortment of behavior change techniques such as managing and monitoring exercise, diet weight and sleep. The existing mobile apps used to promote healthy lifestyle are inadequate as they are either fitness trackers or dietary applications. This research focused on designing and developing an integrated fitness and diet prototype which can comprehensively promote health by having an assortment of exercises and diet plans. The functional requirements were determined by carrying out an expert interview on 12 expert, 8 fitness trainers and 4 dieticians selected through cluster and snowballing sampling respectively. The Application was developed using RAD within the Design Science Research Methodology. The prototype was quantitatively evaluated using a usability test on 197 fitness enthusiasts who were selected using systematic sampling from 8 gyms located in different areas within Nairobi. Links to the APK file and Google Form questionnaire were shared to the prospective subjects through whatsapp groups of the selected gyms. 81% of the users found FitLink to be pleasant to use with over 60% indicating perceived acceptance of the application. Based upon the respondents feedback, Conclusion: The results demonstrate that a functionality unified mobile application is effective in providing all-inclusive information on exercise and diet to the fitness enthusiasts and by extension the larger population that may require to change their health lifestyle.

Project Title: A Toolkit Approach for ISO/IEC 27001 Standard Implementation, Case of SMEs in Nairobi, Kenya

Abstract: Information security has become an important aspect in today’s business environment, where all operations are technology centered. Over the years lots of effort has been put to ensure that organizations manage information security in a standardized manner. There are several frameworks and standards such as COBIT, ITIL and ISO/IEC 27001 that have been proposed for this purpose. In this work, the focus was on ISO/IEC 27001 which is an international standard that provides specification for an Information Security Management System (ISMS). The standard is designed to assist large and small enterprises to manage their information security processes in line with international best practice. Small and Medium-sized Enterprises (SMEs) usually find it difficult to comprehensively implement the prescriptive requirements of the standard. This study proposes a toolkit approach in helping SMEs implement the requirements of the standard. It proposes and develops an ISO/IEC 27001 information security toolkit as a prototype for guiding organizations in implementing information security controls. Apart from toolkit design and implementation, the study also assesses the toolkit and its usability. Results indicated that majority of SMEs would embrace the toolkit and that it can be of great importance in guiding them implement controls of the standard. Furthermore, the study found out that with further enhancement of the toolkit features, to incorporate all aspects of ISO 27001 standard, the toolkit can be used for both large enterprises and small enterprises in implementing the standard requirements.

Project Title: Cyber Security Readiness Assessment Model in Kenya’s Higher Learning Institutions: A Case of the University of Nairobi

Cyber security is an important pillar to effective operations on a network infrastructure integrated with information and communications technology. The world today has rapidly embraced the internet whilst personal, social and professional lives have gone digital. While technology and innovation continue to modernize the way we do things, securing the systems and infrastructure lags behind. Due to the ever evolving and growing advancement in digital crime, the computer and network security becomes a fundamental issue. Information technology networks security objectives seek to maintain confidentiality, integrity and availability. Given the openness and extent of cyberspace, it is possible for offenders to conduct covert attacks and exploit vulnerability in systems. In order to secure the systems, higher learning institutions should conduct extensive direct examination in order to assess the cyber security readiness levels. Research shows that there exist various models which have been developed for cyber security readiness assessment; however, they are customized for developed countries whose cyber networks are much more advanced and may not be directly applicable in the case of developing economies. Therefore, this study developed a suitable model aimed at assessing the cyber security readiness, targeting information communication technology staff form institutes of higher learning in Kenya. In particular, the study investigated the cyber security readiness of ICT personnel from the University of Nairobi with the objective of determine the factors that influence cyber security readiness, develop a model for cybersecurity readiness assessment and conduct a diagnostic assessment of the ICT Staff in regard to their cyber security readiness.

Project Title: Online Platform for Informal Vocational Learning to Improve the Rate of Knowledge and Skill Acquisition Among Women in Kenya

Abstract: Early pregnancies and restrictive cultural practices constraint the time, mobility and flexibility required for women to access education. Consequently, the overall economic contribution and labour participation among women is exiguous. More research is necessary to investigate the application of technology in promoting women’s access to Technical, Vocational and Entrepreneurship Training (TVET). Such skills provide prospects for career advancement as well as self-reliance in absence of employment. This is especially relevant to Kenya where more than 50% of women have not attained secondary school education level and above. This exploratory study aims at fostering informal vocational learning among women in Makueni County using publicly available content. The methodology adopted was a three-step process comprising of a prestudy to identify the requirements of a self-paced Learning Management System (LMS), development of the LMS as per the identified requirements, and finally, a post-study to evaluate its effectiveness in learning. Stratified purposive sampling was used to identify 30 women with limited prospects for tertiary education for the study. The evaluation for the LMS was done using phone interviews and involved giving the participants access to the LMS for two weeks. Findings of the study showed that 93% of the participants had effectively learned sufficient skills in at least one of the four learning areas provided in the LMS, and it concludes that offering publicly available content in self-paced LMSs can promote vocational learning among women.

Project Title: Comparing different classification algorithms to predict the adherence to tuberculosis treatment for new cases in Kenya

Abstract: This study determines factors that are associated to non-adherence to tuberculosis treatment in Kenya. In the African Region, over 25% of the tuberculosis deaths occur. Kenya is among the 30 high burden countries accounting for more than 80% of tuberculosis cases in the world. In Kenya, TB is the number five killer. Due to the high cases of TB, WHO established a global plan called End TB Strategy that was aimed at reducing the tuberculosis deaths by 95%. Adherence to TB treatment is a key element to ensuring a successful control TB program, however, not every patient adheres to TB treatment. Non adherence to TB treatment results in the increase in number of deaths, drug resistance by patients, length of illness and disease transmission, which have economic consequences for patients and their families due to loss of income and cost of the health system. A system that tells if a patient will adhere to the tuberculosis treatment or not can help to curb the non-adherence rates.

Project Title: Factors Affecting the Adoption of Thin Client Technology in Public Universities in Nairobi County

Abstract: Advancement in technologies has made virtual classrooms which enables millions of people to access knowledge at their fingertips an essential tech requirement. Public universities in Nairobi county that have embraced thin client technology adoption which is the main focus in this study, work to establish cost maximization, energy-saving opportunities, solve a wide range of technical and logistical problems also helps top management with strategic decisions, value‐added customer service, data security.

Project Title: Early warning system for flash flood management case of Hell’s Gate National park

Abstract: Natural disasters, like flood, flash floods, earthquakes have led to destructions of properties, loss of lives around the world. Hells gate national park in recent years has experiences destructive flash floods, these situations prompted the development of flash flood early warning systems technology has been useful in mitigation of floods which causes deaths and destruction, notification and prior knowledge of the occurrence of the flash flood is critical in saving lives. The use of modern GSM and wireless technology is key in passing information, risks avoidance, preparedness, mitigation and response. Data collection of the study was based on an assessment of how early warning systems can be used effectively in ensuring safety of tourist visiting the park and specifically the gorge, and therefore primary data from the parties involved in this matter is crucial. However, secondary data from relevant publications and internet has been used to support various augments in the studies. The early warning system will be effective tool in early warning system.

Project Title: Enhanced Mobile Banking Security: Implementing Transaction Authorization mechanism via USSD Push.

Abstract: Mobile initiated financial transactions need to be authenticated. This is a mandatory requirement since it serves as a security step or mechanism against non-repudiation. This is true for Mobile Banking customers in Kenya. The stage of protection for a given authentication scheme relies upon on characteristic combination, authentication channel, credential storage, and encryption. A range of researches had been performed on mobile banking authentication and their stage of protection. Research has proven challenges related to single factor or two factor authentication schemes. However, there are inadequate studies on authentication schemes that mixes different factors of
authentications for secure and efficient mobile banking transactions.
The goal of the research was to explore challenges of using PIN as the only factor of authentication and further evaluate the effectiveness of incorporating a combined USSD push and PIN efficient multifactor authentication. Convenience non-probability method was used to identify a subset of the population and Snowball
Sampling used to target a total of 385 respondents. A
total number of 442 responses were received through
online administered questionnaires. The study found
84.4% of the respondents use mobile banking frequently.
That is to say, many times during the daily lives. Further
finding was, the de-facto login method used in mobile
banking applications in Kenya, is via PIN and 69% of
respondents have incurred losses due to compromised
PINs. These descriptive statistics necessitated a need for
secure mobile banking app. Hence a need for multi
factor authentication.
The solution implemented offers remedy to challenges
faced by mobile banking customers in Kenya. This solution
was not entirely user’s PIN dependent but also tied to
other details such as International Mobile Equipment
Identity (IMEI), Mobile Systems International Subscriber
Identity Number (MSISDN), and International Mobile
Subscriber Identity (IMSI) in addition to time bound USSD
push augmented with biometric authentication,
Fingerprint. These attributes were encrypted using BCrypt
Hashing Function in mobile banking applications. The
storage of credentials was in distributed locations in
encrypted format. The architecture employed provided
improved security from cyber-attacks such as: identity
theft, phishing, social engineering, spoofing and man in
the middle attack.
In conclusion, use of USSD push in mobile banking
provides an efficient layer of authentication hence
improved mobile banking security.

Project  Title: Evaluation of User Satisfaction Based on Experience With E-Government Platforms

Abstract: Governments have continued to push to ensure that citizen facing services are automated. That innovation is recognized has the implication that the government has the freehand to innovate on the delivery channels so as to take the service closer to the citizens via online portals, digital payments and self-printing of payment receipts or the requisite documents and e-participation. Many governments around the world have implemented such systems for use not only by their citizenry but indeed all others that may want a particular service from the Government. Difficulties in implementation of technology is an area faced by the developing countries in implementing e-government systems and hence the need to close existing gaps provides opportunity to conduct further studies (Ahmad, et al, 2012). This study set out to investigate citizen satisfaction with e-government websites in Kenya by engaging the end-users to capture their interaction experience primarily on several fronts; the service that they receive from the websites and the interface being the design, layout and navigation and user training through availing policies and user documentation and systems availability. To bring this to fruition, the researcher undertook a survey through use of questionnaires to collect the research data. Descriptive statistics, diagnostic tests, inferential statistics were undertaken on the data collected. SPSS (v25) and SmartPLS were used to undertake the analysis of the data and the results presented. The results of the research findings suggest that user expectation, user interface, and system reference documentation are important factors that influence end-user satisfaction.

Project  Title: Assessment of M&E IT Systems in Community Based Organisations

Abstract: The study was to evaluate the design of M&E IT systems used in Community-Based Organizations (CBOs) and to contribute to data management and program improvement. This assessment was to determine how M&E IT systems comply with established guidelines, identify gaps and developments in the CBO M&E IT system, and understand how to improve program delivery using the CBO M&E IT system module. Similarly, they are confronted with a lack of plans, guidelines & operational documents, inadequate data verification checks, poorly developed data quality systems and training of meal staff and volunteers. Besides, the study proposed that for the community-based institutions should invest in professional training of employees, allocated adequate funds for developing or build effective M&E IT systems and use of standard tools for data collection and reporting.

Title: A Comparative Study of Infrasound Based and Image Based Wireless Sensor Networks for Human Elephant Conflict Mitigation.

Abstract: Wireless sensor networks have been widely used in scientific studies to investigate phenomenon and acquire data from otherwise hostile environments that span over large geographical areas. Elephants have been discovered to produce infrasound: - sound that humans cannot hear, that they use to communicate within the herd. Due to changes in the world today, animals have been observed to move out of their natural habitats and into human settlements in search of food while humans have also encroached into the animals’ natural habitat. Due to this, cases of Human-Elephant conflict are on the rise and need to be addressed. Although many technologies have been implemented to address this issue, they have been reported to be inefficient for early alerting. Two technologies; Image-based and Infrasound-based have been proposed as candidates for application of Wireless Sensor Networks. This research compared the effectiveness and efficiency of an Infrasound Wireless Sensor Network (WSN) and an Image processing based WSN for Human-Elephant Conflict (HEC) mitigation by using a Multi-Agent Simulation. This exploration was done by developing models for each of the technologies; Infrasound based WSN and Image processing based WSN and comparing their outputs after simulating them on a computer under the same variables. Results were collected from the Simulations as graphs which were then exported to spreadsheets for differentiating comparative analysis. Based on the results we concluded that Infrasound based WSN for HEC is a superior technology to an Image processing based WSN for HEC. We recommend an efficiency Matrix based on the area to be covered and the detection efficiency of the sensor before any WSN is deployed while deciding on which option to implement. The research borrows from and extends existing applications of wireless sensor networks and their designs, detection and alert systems, and sound engineering.

Project  Title: Effectiveness of the use of short text messaging service use on people living with HIV/AIDs treatment, prevention and control in Kenya.

Abstract: HIV/AIDs is disease which has caused loose of many lives globally especially in developing countries Kenya amongst them. Kenya is on track to achieve global targets of zero new infections, zero Aids related deaths and zero discrimination by the year 2030. Kenya incorporated use of SMSs in hospitals which was seen as tool for treatment, prevention and control of HIV/AIDs which is long life disease. Despite the use of SMS services,   progress of counteracting HIV/AIDs is slow with estimation of new infections to be 52800 and 28200 deaths of Aids related sickness been reported annually 2017.  This research evaluated effectiveness of short message services (SMS) use on PLHIV for their treatment, prevention and control at SWOP clinic Kenya. Research adapted health belief model to evaluate whether its variables had an influence on effectiveness of short message services use on PLHIV for treatment, prevention and control of Aids. Research design adopted for study was survey research design. Population of study was PLHIV under ART and enrolled for SMS services. Instrument for data gathering was closed ended questionnaires. Sample size was 68. Simple random sampling technique was used. Descriptive and inferential analysis were the methods used for analysis. Study findings concluded 83% of variations in effectiveness of SMS were influenced by perceived susceptibility, perceived severity, perceived benefits, and perceived barriers, cue to action, self-efficacy, confidence, threat, cost and information. The study recommends there is need for public private partnership to be established. Research concluded, MS is an effective tool for treatment prevention and control of HIV/AIDs.

Project Title: Knowledge Management Systems for the Kenya Police Service

Abstract: Police forces around the world engage in a variety of activities. Many of these activities, such as roadblocks, arrests, and increased police presence in problem areas, are reactive, but there are also many proactive measures involved in police work. Both reactive and proactive measures involve a large amount of information, and Kenya police officers need to understand vast amounts of knowledge to carry out their normal duties. This amount of information means that Kenya police officers also have to be proficient knowledge workers. Knowledge Management Systems in Law Enforcement: Technologies and Techniques present the latest research and trends in using knowledge management to aid police activities. The principles and practices presented in this report will help law enforcement professionals find ways to quickly bring vast amounts of knowledge to bear in law enforcement, but will also help prepare the next generation of police officers for a more knowledge intensive field.

Project Title: Evaluation of E-Procurement System Service in Ministry of Health: A Case of Murang’a County government.

Research Supervisor: Christopher A. Moturi

Abstract: Assessing the quality of e-services has now become an interesting area of study. In the past, many issues resulting from system misuse to achieve selfish motives have been recorded. This has resulted to misuse of public funds by specific individuals. The procurement process have had challenges due to many factors among them manipulation of the e-procurement system. This system manipulation has necessitated a lot of public money to get lost in the hands of specific individuals. In the long run, development projects lag behind. This research is intended to assess the e-service quality of E-procurement system in Murano’a County using E-S-QUAL and E-RecS-QUAL scales. E-S-QUAL and E-RecS-QUAL scales are used to measure the quality of e-services. In this research, they were first adjusted and adopted to explore the various dimensions that have an effect on the overall customers’ perceptions of the E-procurement system in place.104 valid questionnaire answers were taken. The questionnaire was structured such that there were 250 structured questions which were given to the various stakeholders of the E-procurement system.
At first, operationalization of the used scales was done. A correlation and factor analysis was done followed by multiple regression analysis to the E-S-QUAL and E-ReS-Qual scales. Based on the results, there is a strong positive correlation between the overall perception of customers and the E-procurement system. The dimension with the greatest impact is the ‘system efficiency, privacy and finally contact. The E-procurement system scored 3.001 from the 5-point scale in performing quality E-services and a score of 2.764 in E-service quality recovery. The biggest limitation of the study was the fact that it is barely 3 years since the launch of system use in county government so the sample size was limited to a given population. The study is of value to policy makers as issues raised will act as a basis of formulating policies that will aid in system improvement. From the results, it implies that the Eprocurement services offered by the system are not satisfactory to the users. Therefore, the management should consistently work on privacy, system responsiveness and system availability issues.

Project Title: A Recommender System for a Local Video Streaming Service

Research Supervisor: Dr. Lawrence Muchemi

Abstract: In the past few years, we have witnessed explosive growth in e-commerce and online streaming services. This has led to ever-growing upload of new content every day, which presents an information overload challenge to users. To filter information from these huge volumes of data and present products that are of interest to users is a very difficult task. To solve this problem, organizations are deploying recommender system to provide suggestions to products that might be of interest to the user. Recommender systems have become prominent in modern web applications such as streaming services and e-commerce due to their capability to personalize user experience by providing items suggestion that the customer will most likely buy. Traditionally this systems have been based on neighborhood techniques and latent factor models. However, these models have not been efficient in utilizing sequential features of historical user transactions. This research designs a deep learning model based on recurrent neural network that processes sequential user-item interaction with data in order to provide personalized recommendations to users based on their past interactions. The outcome of our study show that recurrent neural networks can also achieve good predictive results in recommender systems domain.

Project Title: Factors Influencing Adoption of Cloud Based Human Resource Management System In Kenyan Counties; A Case study of Neri County.

Abstract: A Cloud-based human resource management system (HRMS) presents numerous benefits to an organization including cost reduction, improved efficient, elimination of errors, and optimal utilization of human resource data. The aim of this study was to assess factors that affect the adoption of cloud-based HRMS in county governments resulting in the development of an adoption framework. The specific objectives of the study was to establish the effect of organizational, technological, and external environment factors on adoption of cloud-based HRMS at the Nyeri County Government offices. The study adopted a descriptive case study design with a sample of 160 staff selected from a population of 400 staff working. Findings revealed that Nyeri County Government has made notable progress in terms of adoption of cloud-based HRMS with adoption score of 74.1%. The study recommended that to improve the adoption of cloud-based HRMS, the county government should ensure adequate allocation of resource, improve internet connectivity, and find strategies for coping with a less vibrant economic environment.