http://repository.kemu.ac.ke/handle/123456789/12 Wed, 15 Apr 2026 03:47:13 GMT 2026-04-15T03:47:13Z http://repository.kemu.ac.ke/handle/123456789/2280 An Assessment of Eco-Friendly Digital Records Management Practices for Promoting Environmental Sustainability: A Case Study of the Marsabit County Teaching and Referral Hospital Diba, Bilinga Kosi Environmental sustainability is a growing global concern, driving institutions to adopt eco-friendly practices in their daily operations. This study examined how paperless communication, digital archiving, cloud storage, and e-waste management contribute to sustainability at Marsabit County Teaching & Referral Hospital (MCTRH). Anchored on the Green Information Technology (Green IT) theory, a descriptive survey design was applied. Data were collected from 117 staff members through structured questionnaires and from three top managers via key informant interviews. Random sampling was used for staff, while key informants were purposively selected. Quantitative data were analyzed using descriptive statistics, and qualitative insights were thematically analyzed. Instrument validity was ensured through expert review and pre-testing, and reliability confirmed with Cronbach’s Alpha values above 0.7. The study achieved a 97% response rate. Results indicated that paperless communication is moderately adopted, cutting paper use and costs while supporting sustainability. Digital archiving improved accessibility and reduced physical storage needs, though adoption was inconsistent. Cloud storage enhanced collaboration and accessibility, offering strong sustainability benefits despite infrastructural challenges. E-waste management practices were partial, signaling the need for structured recycling and safe disposal. Other initiatives, including solar energy, green campaigns, and electronic medical records, were evident though unevenly adopted. The study concludes that eco-friendly digital records management significantly fosters environmental sustainability among healthcare. It recommends stronger policies to institutionalize paperless communication, investment in reliable archiving and cloud systems, and robust e-waste management frameworks. These findings contribute to the growing body of knowledge on sustainable healthcare management while offering practical implications for policymakers and administrators aiming to integrate green technologies into health information systems. Wed, 01 Oct 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2280 2025-10-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2239 The Relationship Between User Education and Undergraduate Students’ Perception of University Libraries in Meru County Chepkurui, Kibos Jane In the context of rapid technological advancements, information digitization, and the increasing availability of e-resources, effective user education has become crucial for enabling students to navigate and utilize university library resources. Despite these advancements, the two university libraries in Meru County, Kenya, have experienced suboptimal usage, potentially due to students' perceptions of the library. This study aimed to investigate the impact of user education programs on undergraduate students' perceptions and consequently library usage. The research was guided by objectives focusing on the types of user education programs offered, the extent of student participation, students' perceptions of the quality of these programs, and the barriers affecting user education. The literature was reviewed based on the research objectives. The study employed descriptive statistics and was anchored on the Expectancy-Confirmation Theory by Richard L. Oliver. The study was conducted in Meru County, focusing on two chartered universities: Kenya Methodist University (KeMU), a private university and Meru University of Science and Technology (MUST), a public university. The study employed descriptive statistics. The target population was 6138 first-year undergraduate students enrolled in the academic year 2023/2024. The study employed stratified sampling techniques based on academic schools. The study used Krejcie and Morgan (1970) table to determine the sample size, which was 364 students. The researcher purposively sampled a total of 12 out of 46 library staff. Data was collected from students using questionnaires and interviews for the staff. Pretesting of research instruments was done at Mount Kenya University, Meru Campus. Permission to collect data was sought from the National Commission for Science, Technology, and Innovation (NACOSTI). The computation of descriptive statistics was in the form of mean, mode, median, percentages, and standard deviation. The findings were presented using descriptive tables, figures, and narratives for ease of understanding the results. The findings revealed that library orientation and instruction sessions had high participation rates and were considered effective by the majority of students. Active participation in ongoing user education sessions was moderate, indicating that there was potential for improvement in terms of student involvement. Students generally had positive perceptions of the quality of user education programs. The programs were seen as significant to their educational pursuits, with high satisfaction levels regarding the relevance and adequacy of the resources provided. Barriers to user education included inadequate session time allotment and a lack of current digital resources. Recommendations include increasing the duration and frequency of user education sessions, updating digital resources, and utilizing promotional techniques such as social media for broader outreach. Future research could explore the long-term impact of user education on academic performance. This study contributes new insights into the relationship between user education and library perception, highlighting the importance of tailored educational interventions in enhancing library usage. Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2239 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2157 Social Media Promotion Strategies for Enhancing Student Engagement with Library Services: A Case of Strathmore and Riara University Libraries in Kenya Segel, Winner Naisula University libraries are increasingly adopting social media as a means of promoting their services, yet the effectiveness of these strategies remains underexplored in the Kenyan private higher education context. This study examined how social media promotion strategies enhance library service provision to students at Strathmore and Riara University Libraries. The study focused on four strategies: content creation, user engagement mechanisms, targeted advertising, and gamification. A descriptive mixed-methods design was applied, involving a survey of 300 undergraduate students in Information Technology and Computer Science, of whom 255 responded (85%), and interviews with 36 librarians, of whom 20 participated (55.6%). Questionnaires were used for students, while semi-structured interviews captured insights from librarians. Quantitative data were analyzed using descriptive and inferential statistics, while qualitative data underwent thematic analysis. Reliability was confirmed through Cronbach’s Alpha coefficients above 0.7. The findings revealed that content creation, particularly infographics and regular posts, moderately improved student awareness (mean = 3.15). User engagement remained weak (mean = 2.25), as libraries mainly used platforms for information rather than interaction. Targeted advertising showed minimal impact (mean = 2.88), limited by financial and technical barriers. Gamification emerged as the most effective strategy, with quizzes and contests significantly motivating student participation (mean = 3.42). The study concludes that while social media enhances library visibility, its full potential remains underutilized. Practical recommendations include staff training in digital content creation, investment in interactive tools, and integration of gamification beyond orientations into ongoing library activities. The study contributes to policy and practice by providing an evidence-based framework for optimizing social media strategies in Kenyan university libraries. Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2157 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2156 Deep Learning Approach for Detection and Prediction of Pest Infections On Plants in Greenhouses Sambu, Bridgite Mueni Pest infestations remain a serious threat to greenhouse agriculture ultimately resulting in reduced yields, increased cost of production, and food safety concerns. Traditional pest monitoring strategies are predominantly manual in nature, which can be laborious, time-consuming, and exhibit human error. These downfalls can affect timely action on issues and a positive relationship management with pest monitoring can mean excessive pesticide use with financial and/or environmental impacts. This paper proposed an AI-enabled hybrid deep learning model to automate pest detection and estimate the probability of a pest outbreak. The hybrid model combined Convolutional Neural Networks (CNNs) for spatial (image-based) pest identification with Long Short-Term Memory (LSTM) networks to estimate probabilities of outbreaks based on sequences of environmental ((local) environmental data temperature, humidity, as well as, recorded counts of pests) and pest counts to help improve timing of interventions and proactive pest management solutions. The study made use of both primary and secondary datasets. The primary dataset was collected from three greenhouses in Limuru, Naivasha, and Thika, Kenya as well as high-resolution crop images (48 mega-pixels) representing various pest infections. Local environmental data for temperature, humidity, and pest counts were collected from the greenhouses to use as sequential variables in the LSTM component of the workflow and assist with estimating the accuracy of predicted outbreaks. Secondary datasets of pre-annotated pest images and historical climate records, including PlantVillage and IP102, provided bulked training data that improved models’ robustness and generalizability to unseen datasets. The researcher ensured stratified sampling to capture representation of all greenhouse types, farm sizes, and agro-climatic conditions. As part of preprocessing, all datasets underwent the following steps: image augmentation, noise removal and feature normalization. All model training, including hyperparameter tuning, occurred in a GPU-enabled Google Colab environment, and early stopping was used to avoid overfitting. The hybrid CNN-LSTM model produced a classification accuracy of 94.7%, precision of 93.9%, recall of 93.8%, and F1-score of 93.2%. The time-series predictive component of the LSTM model produced strong predictive performance with a mean absolute error (MAE) of 0.14, and R² value of 0.89. This showed that the environmental and measured sequence data of pests improve outbreak prediction. It was shown that the hybrid model accurately identifies pest infections, as well as predicting their outbreaks, which will contribute to providing early and timely interventions in the pest control system. Both high-resolution image data and measurements of local environments support a scalable and resilient choice for greenhouse pest management - the use of less pesticides, leading to sustainability in agriculture. In conclusion, this study has demonstrated the effectiveness of a hybrid deep learning framework for integrated pest management with operational and financial implications. Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2156 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2155 Development of A Machine Learning-Based Model Using a Decision Tree for Detecting Fake News: Analyzing Techniques for Accurate Content Verification Tomba, Kinkosi Esther With the growing spread of information through social media and online news, identifying fake news has become increasingly important. To explore this issue, the Pew Research Center conducted surveys in the U.S. to examine how adults access news on social platforms, focusing on the behaviors and demographics of users who rely on these channels. This research sought to address a major gap in traditional fake news detection approaches, which were largely manual and lacked the sophistication of advanced machine learning and AI methods. Such conventional techniques struggle to handle the complexity and contextual manipulation of information, where accurate facts can be framed misleadingly. To overcome these shortcomings, the study developed a machine learning–based model to detect fake news by analyzing article content and recognizing patterns of misinformation. It utilized advanced natural language processing (NLP) techniques and supervised learning algorithms. Decision Trees (achieving 99.67% accuracy), Logistic Regression (99.13%), and Random Forest (99.15%). Processes such as tokenization and TF-IDF were applied to train the model on the ISO Fake News dataset, which combined real news from Reuters.com with fake news from unreliable sources flagged by PolitiFact and Wikipedia. Model performance was evaluated using accuracy, precision, recall, and F1-score, all reaching 99.67%, demonstrating exceptional detection capability. This work contributes to the field of machine learning by enhancing NLP methods and improving the effectiveness of fake news detection models. Future research is encouraged to expand datasets, incorporate multiple languages, employ deep learning like RNNs, CNNs, and Transformers (e.g., BERT and RoBERTa) for richer contextual understanding, and establish benchmarks based on real-world case studies.   Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2155 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2154 Integrating Ai Tools for Real-Time Anomaly Detection in Cloud Vpns: A Case of Owncloud. Boyani, Momanyi Zipporah The increasing reliance on cloud-based Virtual Private Networks (VPNs) has significantly improved the security and scalability of digital infrastructures. However, the increasing complexity of these systems introduced new challenges in ensuring their security, particularly in detecting anomalies such as unauthorized access, abnormal traffic, and data breaches in real time. This research addressed the problem of inadequate anomaly detection in the dynamic cloud VPN environments by investigating the integration of Artificial Intelligence (AI) tools to enhance real-time threat detection, using OwnCloud as a case study. The research aimed to identify effective AI models for anomaly detection, develop a real-time AI-based prototype, and evaluate its performance in detecting anomalies within cloud VPN traffic. Both supervised and unsupervised machine learning techniques were explored, including Isolation Forest and Long Short-Term Memory (LSTM) models. Simulated VPN traffic data was generated using Mininet, and Apache Kafka was employed to stream this data in real time to a Spark-based AI detection engine. Anomaly detection outputs were logged and visualized using the Kibana dashboard, while alerts were configured to trigger based on spikes and deviations from normal traffic patterns. The prototype demonstrated the feasibility of AI-based tools in identifying unknown and evolving threats more effectively than traditional signature-based systems. Unlike conventional methods that rely on historical data and static thresholds, the AI-driven system adapted to emerging threat patterns and significantly reduced false positives. The comparative analysis of AI models confirmed that the hybrid (LSTM + Isolation Forest) model was the most effective AI-based approach for anomaly detection in simulated cloud VPN traffic. It not only delivered superior performance metrics but also demonstrated adaptability in real-world scenarios where labeled anomalies are scarce, and encrypted traffic restricts payload inspection. The model recorded the highest Precision of 0.94, Recall of 0.91, F1-Score of 0.92, and Accuracy of 0.93. The developed AI-based prototype system effectively achieved real-time anomaly detection in OwnCloud VPN traffic. Its hybrid architecture, based on LSTM and IF, delivered accurate, timely, and interpretable results, hence validating its potential for integration into real-world cloud security systems. The ROC curve for the real-time anomaly detection prototype revealed exceptional performance, with an AUC score of 0.98 confirming its effectiveness in distinguishing between normal and anomalous traffic. The findings highlighted the potential of AI to improve the responsiveness and accuracy of intrusion detection mechanisms in cloud-based environments. In conclusion, the research successfully demonstrated that AI tools can enhance real-time anomaly detection in the cloud VPNs, offering improved threat response and reduced false alarms. This research will add to the existing knowledge of AI integration to improve the security of cloud VPNs by exploring a case study in real life. It is recommended that future implementations expand on this approach by integrating more advanced deep learning models, refining real-time alert systems, and applying the solution to diverse cloud platforms to further validate scalability and robustness.   Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2154 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2153 Enhancing Diabetes Classification Using A Weighted Ensemble Of Tabnet, Xgboost And Random Forest Obunge, Duncan Ogindo The increasing prevalence of Diabetes mellitus (DM), a leading global health challenge, is significantly impacting the healthcare systems. Accurate and interpretable classification models are crucial in advancing early diagnosis and effective intervention. While traditional machine learning techniques like Extreme Gradient boosting (XGBoost) and RandomForest based models have demonstrated robust classification performance on tabular medical datasets, however, they have continued to face challenge of model interpretability. Deep learning models, like TabNet, cater the two-pronged benefits of feature selection learning and interpretability via attention mechanisms. This study developed a weighted ensemble model that combines TabNet, XGBoost and RandomForest based models to address the trade-off between interpretability and strong performance. The study utilized the Pima Indian Diabetes dataset as secondary data and expert clinical validation. The dataset, contained 768 tuples with 8 features, related to diabetes risk factors. The ensemble assigns optimized weights to the classifications of the three models, drawing on their complementary strengths. The results indicated that the weighted ensemble model outperformed the individual models; while preserving interpretability. The implementation achieved a balanced accuracy of 0.8630 ± 0.0146 (median 0.8350), precision of 0.8163 ± 0.0442 (median 0.8018), recall of 0.9376 ± 0.0341 (median 0.8900), F1 score of 0.8401 ± 0.0110 (median 0.8436), and ROC-AUC score of 0.9026 ± 0.0172 (median 0.9044), while the traditional machine learning models based on XGBoost attained (0.8103 ± 0.0270 (0.8150) balanced accuracy, 0.7888 ± 0.0287 (0.7890) precision) and RandomForest achieved (0.8060 ± 0.0250 (0.8100) balanced accuracy, 0.7451 ± 0.0312 (0.7768) precision) algorithms. Feature importance analysis revealed the top most significant predictors of diabetes based on normalized scores as; glucose level (≈1), followed by age (≈0.458), insulin (≈0.434) and body mass index(BMI) (≈0.13) hence providing valuable clinical insights. This research contributes a novel computational framework that leverages a weighted ensemble learning techniques while preserving model explainability; a critical advancement for healthcare-aligned machine learning systems. This methodological contribution extends beyond diabetes classification to potentially benefit various clinical decision support systems operating on limited-feature medical datasets. Mon, 01 Sep 2025 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2153 2025-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2060 Factors Influencing Commercial Production of Indigenous Poultry in Mandera County, Kenya Mohamed, Mohamud Mohamed Most farmers pursue poultry farming on a small scale as a hobby rather than as a business, thus they are unable to take advantage of the accompanying economic benefits. This is especially true of indigenous poultry. The purpose of this research was to establish the factors that influence commercial indigenous chicken production in Mandera County, Kenya. The specific objectives of the study were; to determine the effects of supplementary feeding, disease control, technical knowledge, market availability and credit access on commercial indigenous poultry production in Mandera County. The study's philosophical foundation was positivist. The study targeted 3971 households in Mandera North, Banisa and Mandera west sub counties in Mandera County, who constituted the population. A sample of 363 respondents, one from each sampled household was selected to represent the population. Primary and secondary data were used. Questionnaires, focus group discussions, and interview schedules were used to gather the main data. Preliminary confirmatory tests for reliability showed that the data collection tool was reliable with a Cranbach Alpha coefficient above 0.7. for most variable constructs. To determine the strength and nature of the independent variables impact on the dependent variable, multivariate regression analysis was carried out. The response rate was at 94% where 344 respondents answered and returned the questionnaire out of 363. The correlation results indicated that the independent variables; credit availability, disease control, market availability, poultry feeding and technical advisory services were strongly and positively correlated to indigenous poultry production as indicated by Pearson Correlation coefficient values of 0.584, 0.612, 0.827, 0.661 and 0.796 respectively. The results of ANOVA revealed that the F calculated value was 736.785, while F critical was 2.76, at a 5% level meaning significant at that level. Tests of hypotheses indicated that the independent variables were significant for poultry feeding (p=0.000), credit availability(p=0.000), disease control (p=0.000), market availability (p=0.000) and technical knowledge (p=0.000). Recommendations were made for policy makers to create an enabling environment for private enterprise to thrive in terms of improving on infrastructure and providing technically qualified support staff. It was further recommended that commercial poultry producers would pool resources to leverage on economies of scale and group dynamics in credit and market access. Mon, 01 Jan 2024 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2060 2024-01-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2048 The Role of Public Library in Improving the Livelihood of Nomadic Communities in Garissa County, Kenya OMAR, ABDI YUSSUF Increasing access to information and knowledge for nomadic communities supports sustainable development and improves the community’s livelihood while establishing active, responsible, and inclusive institutions at all levels. This study investigated the role of public libraries in enhancing the livelihoods of nomadic communities in Garissa County. The objectives were to determine the information needs, examine the information sources and services provided by public libraries, and evaluate strategies that enhanced community participation in public library programs and livelihood improvement in Garissa County. The research was anchored by the Wilson Model of information seeking behaviour, social inclusion, innovation theory, and community development theory. Descriptive survey research design was adopted. The target population was 377, including library users, heads of public libraries, and sub-county officers. Data were collected using questionnaires and interview schedules, which were validated and tested for reliability through pretesting in Tana River County. SPSS version 26 was used for data analysis, employing frequency distribution and central tendency measures. Ethical considerations included obtaining a NACOSTI research permit, confidentiality, and academic integrity. The findings revealed that nomadic communities in Garissa County had distinct and pressing information needs to enhance their livelihoods. Sustainable farming practices, effective livestock management, and educational opportunities tailored to the nomadic lifestyle were identified as key areas of interest. By providing tailored resources and programs, public libraries addressed these needs to some extent. However, to meet the unique challenges of nomadic lifestyles effectively, libraries must continue to adapt and expand their services. The study recommends regular needs assessments to identify and address the specific information requirements of nomadic communities. Furthermore, public libraries should collaborate with community leaders to ensure that their collections and services meet the practical and cultural needs of the nomadic community, fostering greater engagement and livelihood improvement. Sun, 01 Sep 2024 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2048 2024-09-01T00:00:00Z http://repository.kemu.ac.ke/handle/123456789/2045 Influence of Financial Investment, Utilization Patterns, Perceived Value and Support Services on Maximizing Electronic Resources Usage at Technical University of Mombasa and Kenya Methodist University KOCHUMBA, TERESIA ATIENO Academic libraries at Technical University of Mombasa (TUM) and Kenya Methodist University (KeMU) invest heavily in electronic resources to foster academic excellence and enhance research quality. This study assessed the influence of financial investment, utilization patterns, perceived value, and support services in maximizing electronic resources usage at Technical University of Mombasa and Kenya Methodist University with a view to optimize resource allocation and further enhance the benefits of electronic resources. Both universities had increased their financial investments on electronic resources, and introduced digital libraries, social media engagement/ marketing to promote the full maximization of these resources. Despite these efforts, both institutions continued to face challenges such as the misallocation of financial investments, limited usage patterns, low perceived value, and ineffective support services. Additionally, TUM's electronic resource usage was only 32% while KeMU’s was 57% in 2022/2023 academic year. This study aimed to evaluate the impact of the strategies implemented by TUM and KeMU in maximizing electronic resource usage. It focused on key objectives, including identifying financial investments, determining utilization patterns, assessing perceived value, and establishing the support services in place to enhance electronic resource usage at both institutions. The findings were intended to guide future budget allocations and customize support services to improve resource utilization, perceived value, and ultimately, the maximization of electronic resources at TUM and KeMU. Guided by Edward Freeman’s 1984 Stakeholder Theory and employing a descriptive survey research design with a mixed-methods approach, this study surveyed 426 individuals from a target population of 23,039. The sample included 220 undergraduate and 80 postgraduate students selected through stratified random sampling, 120 faculty members selected through stratified proportional sampling, and 6 library staff selected using purposive sampling. Data was collected through questionnaires and interviews, with pretesting conducted at the University of Nairobi-Mombasa campus to ensure the reliability and validity of the research instruments. Data was analyzed using SPSS for quantitative measures and thematic analysis for qualitative insights. The study achieved a response rate of 91.3% for undergraduate, postgraduate, and faculty members, and 100% for library staff. Findings indicated that 83.3% of library staff were satisfied with budget allocations for electronic resources. Additionally, 75.8% of users engaged actively with these resources, with audiovisual materials being the most utilized and mobile phones being the preferred access device. User satisfaction was reported at 73.1%, and 96.4% expressed contentment with support services. The study highlighted that additional financing and user engagement in resource acquisition improved perceived value and utilization. However, 24.2% of respondents never used the resources, and 23.7% reported dissatisfaction. Both universities had implemented support services to enhance resource use, with user support services, internet access, and feedback mechanisms being the most frequently utilized. The study concluded that financial investment significantly impacts perceived value, while support services influences both perceived value and utilization rates. Frequent utilization, supported by effective services, greatly improves perceived value and maximizes the benefits of electronic resources. Recommendations include engaging users in the acquisition process, enhancing capacity building for librarians, and conducting regular user assessments to tailor support services. These steps will improve perceived value, utilization rates, and overall resource maximization. These findings should guide both libraries in developing policies and strategies to further enhance the return on investment from electronic resources at TUM and KeMU. Sun, 01 Sep 2024 00:00:00 GMT http://repository.kemu.ac.ke/handle/123456789/2045 2024-09-01T00:00:00Z