Journal Articles

Permanent URI for this collection

http://192.168.18.118:4000/handle/20.500.12661/350

Browse

Now showing 1 - 20 of 244

3D printing: developing countries perspectives
(Foundation of Computer Science, 2014) Ishengoma, Fredrick R.; Mtaho, Adam B.
For the past decade, 3D printing (3DP) has become popular due to availability of low-cost 3D printers such as RepRap and Fab@Home; and better software, which offers a broad range of manufacturing platform that enables users to create customizable products. 3DP offers everybody with the power to convert a digital design into a three dimensional physical object. While the application of 3DP in developing countries is still at an early stage, the technology application promises vast solutions to existing problems. This paper presents a critical review of the current state of art of 3DP with a particular focus on developing countries. Moreover, it discusses the challenges, opportunities and future insights of 3DP in developing countries. This paper will serve as a basis for discussion and further research on this area.
A deep feature interaction and fusion model for fake review detection: Advocating heterogeneous graph convolutional network
(Elsevier BV, 2024) Duma, R. A.; Niu, Z.; Nyamawe, A. S.; Manjotho, A. A.
In contemporary real-world scenarios, opinion spammers are hired to fabricate reviews that unfairly promote or demote particular products or services for personal gain. Although considerable attention has been devoted to addressing the problem, existing approaches often overlook the heterogeneous nature of reviewer–product interactions. Specifically, the correlation between review text (comments) and overall ratings, which provides various latent rich information to expose fake reviews, remains inadequately explored. Current methodologies focus on limited interactions, such as reviewer–review, product–review, or reviewer–product interactions, while neglecting significant aspects like reviewer–review–product and reviewer–rating–product interactions, leading to inadequate classifier performance. Motivated by this observation, this study proposes a novel Deep Feature Interaction and Fusion Model (DFIFM) whose ideas are five-folds: (a) constructing a reviewer–product interaction bipartite graph that represents heterogeneous feature node interactions through review text and overall rating values; (b) recognizing the existing mutual interactive relationship between review text and overall rating features, we construct a unified GCN to gain additional insights into feature relationships and capture mutual heterogeneous interactions between nodes; (c) to handle the encoding of unstructured review text features as edge attributes, we adopt a convolutional neural network (CNN); (d) attention mechanisms and fusion techniques are employed to capture interdependencies among reviewer–product latent features; and (e) a Multilayer Perceptron (MLP) utilizes the resulting latent feature representation for review classification. Experimental results on three publicly available datasets demonstrate its superiority over state-of-the-art baselines.
A deep feature interaction and fusion model for fake review detection: Advocating heterogeneous graph convolutional network
(Elsevier BV, 2024) Duma, R. A.; Niu, Z.; Nyamawe, A. S.; Manjotho, A. A.
In contemporary real-world scenarios, opinion spammers are hired to fabricate reviews that unfairly promote or demote particular products or services for personal gain. Although considerable attention has been devoted to addressing the problem, existing approaches often overlook the heterogeneous nature of reviewer–product interactions. Specifically, the correlation between review text (comments) and overall ratings, which provides various latent rich information to expose fake reviews, remains inadequately explored. Current methodologies focus on limited interactions, such as reviewer–review, product–review, or reviewer–product interactions, while neglecting significant aspects like reviewer–review–product and reviewer–rating–product interactions, leading to inadequate classifier performance. Motivated by this observation, this study proposes a novel Deep Feature Interaction and Fusion Model (DFIFM) whose ideas are five-folds: (a) constructing a reviewer–product interaction bipartite graph that represents heterogeneous feature node interactions through review text and overall rating values; (b) recognizing the existing mutual interactive relationship between review text and overall rating features, we construct a unified GCN to gain additional insights into feature relationships and capture mutual heterogeneous interactions between nodes; (c) to handle the encoding of unstructured review text features as edge attributes, we adopt a convolutional neural network (CNN); (d) attention mechanisms and fusion techniques are employed to capture interdependencies among reviewer–product latent features; and (e) a Multilayer Perceptron (MLP) utilizes the resulting latent feature representation for review classification. Experimental results on three publicly available datasets demonstrate its superiority over state-of-the-art baselines.
A deep feature interaction and fusion model for fake review detection: Advocating heterogeneous graph convolutional network
(Elsevier BV, 2024) Duma, Ramadhani Ally; Niu, Zhendong; Nyamawe, Ally S.; Manjotho, Ali Asghar
In contemporary real-world scenarios, opinion spammers are hired to fabricate reviews that unfairly promote or demote particular products or services for personal gain. Although considerable attention has been devoted to addressing the problem, existing approaches often overlook the heterogeneous nature of reviewer–product interactions. Specifically, the correlation between review text (comments) and overall ratings, which provides various latent rich information to expose fake reviews, remains inadequately explored. Current methodologies focus on limited interactions, such as reviewer–review, product–review, or reviewer–product interactions, while neglecting significant aspects like reviewer–review–product and reviewer–rating–product interactions, leading to inadequate classifier performance. Motivated by this observation, this study proposes a novel Deep Feature Interaction and Fusion Model (DFIFM) whose ideas are five-folds: (a) constructing a reviewer–product interaction bipartite graph that represents heterogeneous feature node interactions through review text and overall rating values; (b) recognizing the existing mutual interactive relationship between review text and overall rating features, we construct a unified GCN to gain additional insights into feature relationships and capture mutual heterogeneous interactions between nodes; (c) to handle the encoding of unstructured review text features as edge attributes, we adopt a convolutional neural network (CNN); (d) attention mechanisms and fusion techniques are employed to capture interdependencies among reviewer–product latent features; and (e) a Multilayer Perceptron (MLP) utilizes the resulting latent feature representation for review classification. Experimental results on three publicly available datasets demonstrate its superiority over state-of-the-art baselines.
A framework for guiding the design of interactive systems for communication through activities in higher learning institutions
(East African Nature and Science Organization, 2024-01-05) Mwombeki, Anthony; Shidende, Nima; Mselle, Leornard
Human-computer interaction is paramount in the process of designing interactive-computer systems with high capabilities of guaranteeing users’ satisfaction. However, lack of suitable frameworks complicates their effective design. This study investigated suitable frameworks capable of guiding appropriate design of interactive systems for communication through activities (ISCA) in higher learning institutions in Tanzania. These systems are necessary for augmenting communication thereby overcoming persistent challenges related to their reliance on face-to-face, phone-based, and social networking websites communications. This study used a thorough user centred design approach where a framework for human activity design-centred (HADC) was employed. Authors qualitatively analysed data on communication and interaction issues from participants involved. Design science research was then combined with activity theory to develop a four-phased methodology which was used to design an intended human activity interactive communication (HAIC) framework. The findings confirmed that interactions design techniques based on human-activities undertaken are fundamental in designing communication frameworks capable of guiding suitable design of interactive systems in a particular setting. In the context of this study, and with ISCA need, the obtained HAIC framework was found to be appropriate. This study found out that the current user-centred design approach does not explain precisely how designers can employ activity-based interaction design techniques into the design process. Thus, the study contributes to the literature on users’ involvement in interactive systems’ design through the HADC framework provided
A Framework for Guiding the Design of Interactive Systems for Communication through Activities in Higher Learning Institutions
(East African Nature and Science Organization, 2024) Mwombeki, Anthony; Shidende, Nima; Mselle, Leornard
Human-computer interaction is paramount in the process of designing interactive-computer systems with high capabilities of guaranteeing users’ satisfaction. However, lack of suitable frameworks complicates their effective design. This study investigated suitable frameworks capable of guiding appropriate design of interactive systems for communication through activities (ISCA) in higher learning institutions in Tanzania. These systems are necessary for augmenting communication thereby overcoming persistent challenges related to their reliance on face-to-face, phone-based, and social networking websites communications. This study used a thorough user centred design approach where a framework for human activity design-centred (HADC) was employed. Authors qualitatively analysed data on communication and interaction issues from participants involved. Design science research was then combined with activity theory to develop a four-phased methodology which was used to design an intended human activity interactive communication (HAIC) framework. The findings confirmed that interactions design techniques based on human-activities undertaken are fundamental in designing communication frameworks capable of guiding suitable design of interactive systems in a particular setting. In the context of this study, and with ISCA need, the obtained HAIC framework was found to be appropriate. This study found out that the current user-centred design approach does not explain precisely how designers can employ activity-based interaction design techniques into the design process. Thus, the study contributes to the literature on users’ involvement in interactive systems’ design through the HADC framework provided
A framework for security improvement on usage of mobile money application based on iris biometric authentication method
(Informa UK Limited, 2024) Rashidi, Florence U.; Mohsini, Mustafa H.; Mega, Bakari
Offering transactions through mobile devices has many advantages, such as cashless payments, lower transaction costs, and provide employment opportunities. However, it introduces access security challenges that must be dealt with, which may allow unauthorized access, resulting in theft. This work proposes a framework to improve security on the usage of Mobile Money Services (MMS) by using two-factor authentication (2FA) of PIN and iris biometric authentication method (IRBAM). The rapid application development (RAD) approach was used to develop mobile money applications based on the proposed framework. The proposed framework will improve the security of accessing MMS.
A review on deep learning aided pilot decontamination in massive MIMO
(Informa UK Limited, 2024-02-29) Victor, Crallet M.; Mvuma, Alloys N.; Mrutu, Salehe I.
In multi-antenna systems, advanced techniques such as massive multiple-input multiple-output (MIMO), beamforming, and beam selection depend heavily on the accurate acquisition of the channel state. However, pilot contamination (PC) can be a major source of interference which degrades they are performance. Moreover, the severity of PC increases as more pilots are reused between users in the wireless systems. Researchers have shown that PC can be mitigated by using deep learning (DL) approaches. Nevertheless, when minimizing PC, the examination that identifies the applications and factors that distinguish these DL approaches is still limited. This paper reviews these DL approaches and the improvements needed to enhance their performance. Simulation results confirm that DL networks that learn to predict the channels directly have superior performance under PC.
A review on deep learning aided pilot decontamination in massive MIMO
(Informa UK Limited, 2024) Victor, Crallet M.; Mvuma, Alloys N.; Mrutu, Salehe I.
In multi-antenna systems, advanced techniques such as massive multiple-input multiple-output (MIMO), beamforming, and beam selection depend heavily on the accurate acquisition of the channel state. However, pilot contamination (PC) can be a major source of interference which degrades they are performance. Moreover, the severity of PC increases as more pilots are reused between users in the wireless systems. Researchers have shown that PC can be mitigated by using deep learning (DL) approaches. Nevertheless, when minimizing PC, the examination that identifies the applications and factors that distinguish these DL approaches is still limited. This paper reviews these DL approaches and the improvements needed to enhance their performance. Simulation results confirm that DL networks that learn to predict the channels directly have superior performance under PC.
Accessibility and utilization of mobile phones for governance of water resources in the Lake Victoria basin: constraints and opportunities in Tanzania
(Academic Journals, 2015) Mongi, H. J.; Mvuma, A. N.; Kucel, S.; Tenge, A. J.; Gabriel, M.
Mobile phones are the worlds fastest growing technologies in terms of coverage and adoption. This trend makes them a desirable information and communication technology (ICT) platform for addressing the striking challenges of water resources governance at all levels. Some of remarkable challenges in Lake Victoria Basin (LVB) that have negative impact on the water resources are rapid population growth, unfavorable climate change and variability, and non-sustainable urbanization. Mobile phones can contribute to addressing these challenges by addressing three key principles of governance: transparency in a manner that the resources are utilized by increasing population; accountability in both supply and demand sides of water resources for household, industry and agriculture; and participation of key stakeholders in policy formulation and decision making across temporal and spatial scales. This paper analyses the accessibility and utilization of mobile phones among various actors in the governance of water resources within LVB with the aim of identifying opportunities and constraints at community (micro), local government (meso) and national lake/river basin (macro) scale. To achieve this, three data collection sites were identified in Tanzania: Mwanza, Kagera and Mara. Data were gathered from key stakeholders through five methods: key informant interview, actors panel discussion, observation, literature reviews and questionnaire administration. Qualitative data were subjected to contents analyses while empirical data were analyzed using Statistical Package for Social Sciences (SPSS). The analyses provided information on the constraints in access and utilization as well as opportunities availed by mobile phones. Furthermore, their potential for enhancing transparency, accountability and stakeholders participation were summarized and discussed. Conclusion drawn from this study is that there is variation in access and utilization of mobile phones across scales of water resources governance in LVB that offer both opportunities and constraints. Integrated and innovative system that bridges the gap by capitalizing on the opportunities and minimizing the constraints could be an option.
Accessibility and utilization of mobile phones for governance of water resources in the Lake Victoria basin: Constraints and opportunities in Tanzania
(AJOL, 2015) Mongi, H. J; Mvuma, A. N.; Kucel, S.; Tenge, A. J; Gabriel, M.
Mobile phones are the world’s fastest growing technologies in terms of coverage and adoption. This trend makes them a desirable information and communication technology (ICT) platform for addressing the striking challenges of water resources governance at all levels. Some of remarkable challenges in Lake Victoria Basin (LVB) that have negative impact on the water resources are rapid population growth, unfavorable climate change and variability, and non-sustainable urbanization. Mobile phones can contribute to addressing these challenges by addressing three key principles of governance: transparency in a manner that the resources are utilized by increasing population; accountability in both supply and demand sides of water resources for household, industry and agriculture; and participation of key stakeholders in policy formulation and decision making across temporal and spatial scales. This paper analyses the accessibility and utilization of mobile phones among various actors in the governance of water resources within LVB with the aim of identifying opportunities and constraints at community (micro), local government (meso) and national lake/river basin (macro) scale. To achieve this, three data collection sites were identified in Tanzania: Mwanza, Kagera and Mara. Data were gathered from key stakeholders through five methods: key informant interview, actors’ panel discussion, observation, literature reviews and questionnaire administration. Qualitative data were subjected to contents analyses while empirical data were analyzed using Statistical Package for Social Sciences (SPSS). The analyses provided information on the constraints in access and utilization as well as opportunities availed by mobile phones. Furthermore, their potential for enhancing transparency, accountability and stakeholders’ participation were summarized and discussed. Conclusion drawn from this study is that there is variation in access and utilization of mobile phones across scales of water resources governance in LVB that offer both opportunities and constraints. Integrated and innovative system that bridges the gap by capitalizing on the opportunities and minimizing the constraints could be an option.
Accessible urban by modeling the explicit data using fuzzy logic
(ACM, 2017) Daniel Sinkonde; Leonard Mselle; Nima Shidende; Sara Comai; Matteo Matteucci
The paper explores the level of accessibility for people with specific mobility problems in an urban setting. The work is part on the polisocial Maps for Easy Path (MEP) project [1]. The project aiming at developing a set of innovative tools and solutions for the enrichment of geographical maps with information about accessibility of urban pedestrian areas for people with mobility challenges. This paper builds a FuzzyMEP predictive model that accounts for uncertainty in the real world using fuzzy logic to predict the quality and the condition of the accessibility of a path. The FuzzyMEP model has been developed for mapping two explicit data sets, namely the coordinates where a picture of an obstacle on the sidewalk has been taken and the comment. The current urban city accessibility tool uses spatial data and focuses mainly on the accessibility of a path. We present a novel MEP tool by using a fuzzy logic model for pedestrian path prediction of the condition of the accessibility of a path. In this paper, the FuzzyMEP model has been developed by applying a fuzzy logic rule which can be adapted to other contexts.
Adaptive synchronization in complex networks with non-delay and variable delay couplings via pinning control
(Elsevier, 2014) Liang, Yi; Wang, Xingyuan; Eustace, Justine
The adaptive pinning synchronization in complex networks with non-delay and variable delay couplings is investigated in this letter. For both reducible and irreducible coupling matrixes, novel and practical synchronization criteria are obtained by Lasalle's invariance principle, and sufficient conditions for achieving the synchronization to require the minimum number of pinning nodes are derived. Moreover, the method for calculating the number of pinning nodes is given using the decreasing law of maximum eigenvalues of the principal submatrixes. At last, simulation examples are given to verify effectiveness of the proposed pinning synchronization scheme.
Adaptive weighted recurrence graphs for appliance recognition in non-intrusive load monitoring
(IEEE, 2021) Faustine, Anthony; Pereira, Lucas; Klemenjak, Christoph
To this day, hyperparameter tuning remains a cumbersome task in Non-Intrusive Load Monitoring (NILM) research, as researchers and practitioners are forced to invest a considerable amount of time in this task. This paper proposes adaptive weighted recurrence graph blocks (AWRG) for appliance feature representation in event-based NILM. An AWRG block can be combined with traditional deep neural network architectures such as Convolutional Neural Networks for appliance recognition. Our approach transforms one cycle per activation current into an weighted recurrence graph and treats the associated hyper-parameters as learn-able parameters. We evaluate our technique on two energy datasets, the industrial dataset LILACD and the residential PLAID dataset. The outcome of our experiments shows that transforming current waveforms into weighted recurrence graphs provides a better feature representation and thus, improved classification results. It is concluded that our approach can guarantee uniqueness of appliance features, leading to enhanced generalisation abilities when compared to the widely researched V-I image features. Furthermore, we show that the initialisation parameters of the AWRG's have a significant impact on the performance and training convergence.
Addressing land degradation in Tanzania: Contemporary issues related to policies and strategies
(Elsevier, 2012) Mongi, Hector
Several attempts have been made since the early years of independence of Tanzania to address land degradation. Policies and strategies were formulated that emphasized on the importance of land conservation in achieving economic growth. However, the context in which the earlier strategies were designed and implemented led to further land degradation. Failure of these interventions was partly because policies which backed them remained sectoral and most of them adopted the top-down approach with very little involvement of communities. The United Nations Conference on Environment and Development (UNCED) of 1992 in Rio de Janeiro created a new chapter in the country by insisting on the importance of sustainable development. The new policies and strategies were formulated with intentions to correct the past mistakes. This paper has assessed the extent to which Tanzania’s post-UNCED policies and strategies address the recent challenges of climate change on land degradation as reported by the IPCC in 2007. It involved review of selected policies, strategies and related literature on land degradation and climate change. The National Strategy for Growth and Reduction of Poverty (NSGRP) and two other sectoral strategies were reviewed in addition to the environmental, agricultural and water policies. The study finds that significant short and long-term measures to combat the negative effects of land use practices on the land resource are addressed. However, constant review of the policies and strategies is required to ensure that emerging challenges are addressed as part of the overall efforts to combat impacts of global climate change
Against the “Hello World”
(Foundation of Computer Science, 2014) Mselle, Leonard J.; Kondo, Tabu S.
Computer programming is a "two-way thinking process.” The programmer must think and implant his/her thought in the computer in the form of code. In return, the computer must think like the programmer in the way of output. Compilation is the only initial accurate way of confirming that the programmer and the computer are thinking the same way. In case of novice programmers, the compiler feedback does not suffice the need. In some cases it is a source of confusion and despair. To add to this complexity, the way initial programming is taught and the way programming materials are presented to learners goes contrary to the "two-way thinking.” There is a need for another (mediating) language between the compiler and the novice programmer. In this paper, the traditional practice of introducing programming lessons through programs that display a message such as "Hello World" or any other message is debunked. A new visualization approach through Memory Transfer Language (MTL) is proposed. It is proved that MTL is a language to learn programming whereby students are able to employ hands-on, minds-on and "two-way-thinking" approach to develop coding skills.
Age of teachers Vs ICT use in Tanzanian secondary schools: A case of Dodoma municipality
(Foundation of Computer Science, 2015) Mazoya, Adam; Ismail, Abbas; Manyilizu, Majuto
The use of Information and Communication Technology (ICT) in teaching and learning has been the main target for public and private schools in Tanzania since the beginning of the 21st century. Since then, however, the achievement of using ICT on teaching and learning has been relatively unsatisfactory due to several reasons including incompetence and poor understanding of teachers on ICT. This study intends to understand ICT literacy level, use of ICT tools in delivering contents, and readiness to adopt new methods of teaching and learning using ICT among secondary schools teachers with respect to their ages and level of education. The study was carried out in Dodoma municipality in central Tanzania for both government and private secondary schools. Questionnaires facilitated the survey methodology by involving sixteen (16) government and private secondary schools with study sample of 231 teachers. The results show that there is a positive relationship on age of teachers towards their ICT literacy. Furthermore, the study reveals that there is a significant difference between age of teachers and their highest level of education. Contrary, the findings reveal that; age is not a factor on teachers’ use of ICT tools in delivering contents, and the readiness of teachers in adopting new methods of teaching and learning using ICT. Such results provide school administration and Tanzania government a clear way on how to use and build capacity of teachers in different ways of using ICT by considering teachers' age.
Algorithm to identify location of incidents in public buses and respective nearest authority us-ing spherical cosine rule
(IJTEEE, 2013) Hassan, Kilavo; Sam, Anael; Machuve, Dina
Reporting the incidents to authorities in developing countries that can take actions has been a chaotic and slow process. This paper proposes an algorithm for a system that automatically reports events that compromise passengers’ safety to respective authorities by using spherical cosine rule. When an incident such as an accident, over speeding or overloading of passengers occurs the system identifies the closest police station. It does so by calculating the shortest distance between latitudes and longitudes of the place where the event occurred and those of the closest police station. Then it identifies the next police station which the bus will encounter and where the event can be reported and immediate action carried out. The public buses will be equipped with GPS which will generate coordinates and there will be a database which contains the coordinates for the police stations along the main roads.
An empirical study of critical success factors for the adoption of m-government services in Tanzania
(Springer Singapore, 2021) Ishengoma, Fredrick; Mselle, Leonard; Mongi, Hector
For the past decade, the Tanzanian government has started implementing m-government initiatives. However, little is known about the factors surrounding m-government adoption in Tanzania. Consequently, some m-government services have been successfully adopted while others are still struggling (having a low level of adoption). In this chapter, the authors investigate critical success factors (CSFs) that favor the adoption of m-government services from a web analytics point of view. The results show that inspecting the web analytics data from multiple viewpoints and varying levels of detail, gives insights on the CSFs towards the adoption of m-government services. The findings suggest that perceived usefulness, user needs, and usability favor the adoption of one m-government service over the other. Moreover, factors like the loading time of the service, the number of requests, and bounce rate seem not to have an effect.
An ensemble design of a cash crops-warehouse receipt system (WRS) based on blockchain smart contracts
(Emerald, 2022) Shao, Deo; Kombe, Cleverence; Saxena, Stuti
Purpose Warehouse receipt system (WRS) schemes aim to revitalize the market of cash crops, particularly those of the smallholder farmers. However, the existing government schemes for the smallholder farmers have not been much successful on account of issues related with transparency and trust deficit, thereby limiting farmers' access to credit for their farming activities as channeled via the formal financial institutions. Conceding that blockchain is an emerging technology, this study aims to theoretically posit the potential to optimize WRS in real time to bolster farmers' trust and privacy. Design/methodology/approach The research bases itself on the design science research (DSR) approach. Using the DSR approach backed by the literature review and evolutionary prototyping, a blockchain-based WRS has been advanced in the study. Findings This study provides a design of a digital WRS, which serves as a proof of concept of blockchain technology (BCT) applications in improving transparency in the WRS in the agricultural sector. The findings from this study present the possibility of leveraging blockchain smart contracts to improve the linkages between the WRS stakeholders to promote trust and transparency that would eventually unlock the potentials of WRS schemes. Furthermore, the findings inform decision-makers and practitioners to prioritize the emerging technologies in revamping the classical WRS. Research limitations/implications The study leaves research pointers for implementing BCT in agricultural supply chain and furthering the findings from the present study with an empirical agenda. Originality/value Whilst research on the BCT applications across myriad sectors abounds, research on the BCT applications to address societal needs is conspicuously limited. Furthermore, research on the practical implementation of the BCT in real-world scenarios is few and far between. The present study seeks to plug these gaps by underlining the application and utility of blockchain in WRS. The study makes a significant contribution to the theory and practice of the agrarian supply chain management, specifically the WRS, by invoking design knowledge obtained through an iterative design process besides serving as an edifice for the integration of emerging technologies, notably BCT in the agricultural sector.

Browse

Browsing Journal Articles by Title

Results Per Page

Sort Options