OpenUCT :: Browsing by Subject "computer science"

Browsing by Subject "computer science"

Now showing 1 - 20 of 23

Open Access
A mobile application promoting good contact lens practices
(2022) Naidoo, Terushka; Berman, Sonia
Contact lens complications pose an ongoing problem for both optometrists and contact lens wearers. Most of these complications are due to noncompliance to good care practices. Education is the first step to ensuring compliance. If good habits are created on commencement of wear, patients are more likely to continue with these habits or practices. The key however, is maintenance and building on this education, as we cannot expect patients to remember all the information given to them initially. Telemedicine is rapidly becoming a wide reaching and convenient way to provide services and support to patients. The aim of this study was to create a mobile application to provide contact lens wearers with knowledge and assistance to empower them to take good care of their eyes and lenses. A mobile application was built for the study with three main features: a lens change reminder, an information feature, and a diagnosis facility to aid contact lens wearers when they encounter any problems. A PDF version of the application was also created with the latter two features; a secondary aim was to compare its success with that of the mobile application. After receiving ethical clearance for the study, lens wearers who signed the Informed Consent form, were surveyed about their symptoms, knowledge and habits in relation to contact lenses and their eyes. After being divided into two groups, they were either given the mobile application or the PDF document to use. They were subsequently given a second survey to determine if there were any changes to symptoms, habits and knowledge. They were also questioned about the value and effectiveness of the application and the PDF. Although, the results of habit changes were inconclusive, there was a decrease in symptoms after using both the app and the PDF. Both were well received and the majority of participants reported that they would recommended them to others. The mobile application was used more frequently than the PDF, led to a slightly better improvement in knowledge, and scored slightly better in its user evaluation, compared to the PDF.
Open Access
A new connectivity strategy for wireless mesh networks using dynamic spectrum access
(2021) Maliwatu, Richard; Johnson, David; Densmore, Melissa
The introduction of Dynamic Spectrum Access (DSA) marked an important juncture in the evolution of wireless networks. DSA is a spectrum assignment paradigm where devices are able to make real-time adjustment to their spectrum usage and adapt to changes in their spectral environment to meet performance objectives. DSA allows spectrum to be used more efficiently and may be considered as a viable approach to the ever increasing demand for spectrum in urban areas and the need for coverage extension to unconnected communities. While DSA can be applied to any spectrum band, the initial focus has been in the Ultra-High Frequency (UHF) band traditionally used for television broadcast because the band is lightly occupied and also happens to be ideal spectrum for sparsely populated rural areas. Wireless access in general is said to offer the most hope in extending connectivity to rural and unconnected peri-urban communities. Wireless Mesh Networks (WMN) in particular offer several attractive characteristics such as multi-hopping, ad-hoc networking, capabilities of self-organising and self-healing, hence the focus on WMNs. Motivated by the desire to leverage DSA for mesh networking, this research revisits the aspect of connectivity in WMNs with DSA. The advantages of DSA when combined with mesh networking not only build on the benefits, but also creates additional challenges. The study seeks to address the connectivity challenge across three key dimensions, namely network formation, link metric and multi-link utilisation. To start with, one of the conundrums faced in WMNs with DSA is that the current 802.11s mesh standard provides limited support for DSA, while DSA related standards such as 802.22 provide limited support for mesh networking. This gap in standardisation complicates the integration of DSA in WMNs as several issues are left outside the scope of the applicable standard. This dissertation highlights the inadequacy of the current MAC protocol in ensuring TVWS regulation compliance in multi-hop environments and proposes a logical link MAC sub-layer procedure to fill the gap. A network is considered compliant in this context if each node operates on a channel that it is allowed to use as determined for example, by the spectrum database. Using a combination of prototypical experiments, simulation and numerical analysis, it is shown that the proposed protocol ensures network formation is accomplished in a manner that is compliant with TVWS regulation. Having tackled the compliance problem at the mesh formation level, the next logical step was to explore performance improvement avenues. Considering the importance of routing in WMNs, the study evaluates link characterisation to determine suitable metric for routing purposes. Along this dimension, the research makes two main contributions. Firstly, A-link-metric (Augmented Link Metric) approach for WMN with DSA is proposed. A-link-metric reinforces existing metrics to factor in characteristics of a DSA channel, which is essential to improve the routing protocol's ranking of links for optimal path selection. Secondly, in response to the question of “which one is the suitable metric?”, the Dynamic Path Metric Selection (DPMeS) concept is introduced. The principal idea is to mechanise the routing protocol such that it assesses the network via a distributed probing mechanism and dynamically binds the routing metric. Using DPMeS, a routing metric is selected to match the network type and prevailing conditions, which is vital as each routing metric thrives or recedes in performance depending on the scenario. DPMeS is aimed at unifying the years worth of prior studies on routing metrics in WMNs. Simulation results indicate that A-link-metric achieves up to 83.4 % and 34.6 % performance improvement in terms of throughput and end-to-end delay respectively compared to the corresponding base metric (i.e. non-augmented variant). With DPMeS, the routing protocol is expected to yield better performance consistently compared to the fixed metric approach whose performance fluctuates amid changes in network setup and conditions. By and large, DSA-enabled WMN nodes will require access to some fixed spectrum to fall back on when opportunistic spectrum is unavailable. In the absence of fully functional integrated-chip cognitive radios to enable DSA, the immediate feasible solution for the interim is single hardware platforms fitted with multiple transceivers. This configuration results in multi-band multi-radio node capability that lends itself to a variety of link options in terms of transmit/receive radio functionality. The dissertation reports on the experimental performance evaluation of radios operating in the 5 GHz and UHF-TVWS bands for hybrid back-haul links. It is found that individual radios perform differently depending on the operating parameter settings, namely channel, channel-width and transmission power subject to prevailing environmental (both spectral and topographical) conditions. When aggregated, if the radios' data-rates are approximately equal, there is a throughput and round-trip time performance improvement of 44.5 - 61.8 % and 7.5 - 41.9 % respectively. For hybrid links comprising radios with significantly unequal data-rates, this study proposes an adaptive round-robin (ARR) based algorithm for efficient multilink utilisation. Numerical analysis indicate that ARR provides 75 % throughput improvement. These results indicate that network optimisation overall requires both time and frequency division duplexing. Based on the experimental test results, this dissertation presents a three-layered routing framework for multi-link utilisation. The top layer represents the nodes' logical interface to the WMN while the bottom layer corresponds to the underlying physical wireless network interface cards (WNIC). The middle layer is an abstract and reductive representation of the possible and available transmission, and reception options between node pairs, which depends on the number and type of WNICs. Drawing on the experimental results and insight gained, the study builds criteria towards a mechanism for auto selection of the optimal link option. Overall, this study is anticipated to serve as a springboard to stimulate the adoption and integration of DSA in WMNs, and further development in multi-link utilisation strategies to increase capacity. Ultimately, it is hoped that this contribution will collectively contribute effort towards attaining the global goal of extending connectivity to the unconnected.
Open Access
A user interface for terrain modelling in virtual reality using a head mounted display
(2021) Gwynn, Timothy; Gain, James
The increased commercial availability of virtual reality (VR) devices has resulted in more content being created for virtual environments (VEs). This content creation has mainly taken place using traditional desktop systems but certain applications are now integrating VR into the creation pipeline. Therefore we look at the effectiveness of creating content, specifically designing terrains, for use in immersive environments using VR technology. To do this, we develop a VR interface for terrain creation based on an existing desktop application. The interface incorporates a head-mounted display and 6 degree of freedom controllers. This allows the mapping of user controls to more natural movements compared to the abstract controls in mouse and keyboard based systems. It also means that users can view the terrain in full 3D due to the inherent stereoscopy of the VR display. The interface goes through three iterations of user centred design and testing. This results in paper and low fidelity prototypes being created before the final interface is developed. The performance of this final VR interface is then compared to the desktop interface on which it was based. We carry out user tests to assess the performance of each interface in terms of speed, accuracy and usability. From our results we find that there is no significant difference between the interfaces when it comes to accuracy but that the desktop interface is superior in terms of speed while the VR interface was rated as having higher usability. Some of the possible reasons for these results, such as users preferring the natural interactions offered by the VR interface but not having sufficient training to fully take advantage of it, are discussed. Finally, we conclude that while it was not shown that either interface is clearly superior, there is certainly room for further exploration of this research area. Recommendations for how to incorporate lessons learned during the creation of this dissertation into any further research are also made.
Open Access
An analysis of internet traffic flow in SANReN using active and passive measurements
(2021) Salie, Luqmaan; Chavula, Josiah
National research and education networks (NRENs) in developing regions such as Africa experience various performance issues due to inadequate infrastructure and resources. The South African National Research Network (SANReN) connects universities, research institutions, and oversees science projects such as the Square Kilometre Array. In this study, we conduct active and passive measurements to assess the performance of SANReN and to identify problem areas in the network. Active measurements were done to determine network performance when accessing SANReN internally (using PerfSONAR) and externally (using Speedchecker). We found that SANReN needs to be reinforced in and around Port Elizabeth, Cape Town, and Durban. Universities in these cities had the highest delays and page load times. We found that the network traffic flowing from PE uses circuitous routes to flow to universities in Johannesburg and Pretoria, causing high delays (medians of 25.26 ms to WITS, 25.47 ms to UJ, and 25.95 ms to UNISA) and high page load times (medians of 237.07 ms to WITS, 272.09 ms to UNISA, 280.47 ms to UJ transferring 31594 bytes of data). Using Cape Town as the traffic source resulted in a low median throughput of 5.47 Gbps for internal active measurements. Throughput from Durban to Cape Town was low as well (4.91 Gbps), causing high page load times between these two cities (medians of 350.32 and 305.22 ms from Durban to UCT and UWC respectively). SANReN's passive measurements results show us that there is a ratio of 11.16:1 for download speed to upload speed. We also observe a ratio of 2.29:1 for outbound flows (uploads) to inbound flows. Thus, majority of traffic flows experience low throughput amounts. Based on the test results, we design an SDN model and compare its performance to SANReN. The SDN model's results show that it would increase throughput while decreasing delays and page load times.
Open Access
Artificial neural networks to predict share prices on the Johannesburg stock exchange
(2021) Pyon, Freddie; Moodley, Deshendran
The use of historical data to build models for stock market prediction has been extensively researched. Artificial Neural Networks (ANNs) bring new opportunities for predicting stock markets, and is now one of the leading techniques used for time series and specifically stock market prediction. This study explored the application of ANNs to predict share prices in the banking sector of the South African Johannesburg Stock Exchange (JSE). This study used three companies, i.e. Standard Bank, Nedbank and First National Bank, listed on the JSE as case studies for the use of ANNs for predicting the closing share price for the next day, week and month. Historical share price data from the JSE was integrated with datasets of external factors that influence market. The external factors considered in this study include index data from NASDAQ, the JSE top 40 and all share indexes, the exchange rate and the business cycle indicator (BCI) values from the South African Reserve Bank. Comparative analysis were conducted between traditional regression models and ANN models using the lagged share price as input variable. The effect on prediction performance of using external factors as additional input variables was also explored. The ANN models using only the share price was found in general to perform better than both traditional models and ANNs that used the external factors as additional input variables. The average next month prediction model produced a noticeably smaller prediction error compared to the next week, and next day prediction models for all three banks. The results showed that the introduction of external factors as additional input variables did not lead to an improved prediction performance, over models that used only the share price. This study also highlights the importance of using an appropriate validation method and evaluating model stability for evaluating and developing ANN models for share price prediction in time series data. The results contribute to existing research that indicate that an ANN is more effective than a regression method for predicting banking share prices, and that these predictive models have potential for supporting investment decision making.
Open Access
Babel's Tower: South Africa's Wikipedias
(2019) Graaf, Michael; Densmore, Melissa; Johnson, D
This dissertation is a comparative examination firstly of usage of, and contribution to, the Wikipedias in ten of South Africa’s eleven official languages, and secondly of possible measures to address the situation discovered in the first investigative stage. The historical context is reviewed and it is argued that the number of official languages (and therefore Wikipedias) results from decisions made in colonial and apartheid eras. Public-domain usage and contribution statistics from the Wikimedia Foundation are analysed, revealing poor growth in most cases; possible interventions via both cultural/educational strategies and technological options are reviewed.
Open Access
Co-designing in the real world: managing a multiple stakeholder design process with an NGO
(2018) Brittan Sarah; Suleman, Hussein
Many ICT4D research projects work in collaboration with NGOs in order to meet their development objectives and to increase their interventions’ effectiveness. Herein, aspects of co-design are often applied, where the intention is to include all stakeholders as equal participants in the design process. However, collaborating with NGOs and with users who have reduced access to technology can be challenging. As a result, the ideals of co-design are not easily achieved, due to the vastly differing backgrounds of stakeholders in ICT4D projects. In this thesis, an explicit approach for managing the varying interactions between stakeholders is proposed and described through a case study. The approach was derived from ethnographic action research and participatory design methodologies, led by practical consideration from real-world constraints. The approach is structured around an interactive design process that includes the stakeholder groups in unique ways at each phase of the design process, in order to maximise the contributions in a way that respects their backgrounds and areas of expertise. The proposed approach was evaluated through its implementation in the design of a mobile recordkeeping application, in collaboration with an NGO in Cape Town, South Africa. The NGO comprises of two stakeholder groups: the staff and the micro-entrepreneurs who they empower. The NGO’s focus is to provide training and support over a two-year process to women from low-income communities, by teaching them how to manage their own businesses to become socially and financially independent. The objective of this case study was to design a mobile application that aligned with the recordkeeping curriculum of the NGO and meet the specific requirements and constraints of the target users. Through the implementation of the design approach, the students and staff were able to provide useful and complementary contributions towards the design of the system. A one-month field study of the application with a group of 21 student participants revealed that the system was a suitable solution and appropriately met the needs of the NGO and the end-users. The final evaluation of the stakeholders’ reflections on the design process revealed that it was an appropriate design process to have followed. The results further identified that care must be taken to clarify expectations at each stage of the design process, especially when external factors change, and to frequently communicate with all stakeholders involved. The design approach proposed and employed during this research project, and the unique way that it allowed the stakeholders to contribute, will benefit future ICT4D research projects that are faced with stakeholder groups that vary significantly, where traditional equal participation is not possible.
Open Access
Data replication and update propagation in XML P2P data management systems
(2008) Paulse, Marlon; Berman, S
XML P2P data management systems are P2P systems that use XML as the underlying data format shared between peers in the network. These systems aim to bring the benefits of XML and P2P systems to the distributed data management field. However, P2P systems are known for their lack of central control and high degree of autonomy. Peers may leave the network at any time at will, increasing the risk of data loss. Despite this, most research in XML P2P systems focus on novel and efficient XML indexing and retrieval techniques. Mechanisms for ensuring data availability in XML P2P systems has received comparatively little attention. This project attempts to address this issue. We design an XML P2P data management framework to improve data availability. This framework includes mechanisms for wide-spread data replication, replica location and update propagation. It allows XML documents to be broken down into fragments. By doing so, we aim to reduce the cost of replicating data by distributing smaller XML fragments throughout the network rather than entire documents. To tackle the data replication problem, we propose a suite of selection and placement algorithms that may be interchanged to form a particular replication strategy. To support the placement of replicas anywhere in the network, we use a Fragment Location Catalogue, a global index that maintains the locations of replicas. We also propose a lazy update propagation algorithm to propagate updates to replicas. Experiments show that the data replication algorithms improve data availability in our experimental network environment. We also find that breaking XML documents into smaller pieces and replicating those instead of whole XML documents considerably reduces the replication cost, but at the price of some loss in data availability. For the update propagation tests, we find that the probability that queries return up-to-date results increases, but improvements to the algorithm are necessary to handle environments with high update rates.
Open Access
Detection of HTTPS malware traffic without decryption
(2022) Nyathi, Miranda; Hutchison, Andrew
Each year the world's dependency on the internet grows, especially its functionality relating to critical infrastructure and social connections. More than 80% of internet traffic is encrypted using Transport Layer Security (TLS) protocol, and it is predicted that this number will increase [8]. However, threat actors are also increasingly using the TLS protocol to hide malicious activities such as Command and Control, loading malware into a network, and exfiltration of sensitive data. The use of TLS by threat actors poses a challenge to security professionals as traditional techniques used in the detection of HTTP malware cannot be applied in detecting Hypertext Transfer Protocol Secure (HTTPS) encrypted malware. To manage this, companies are using a traditional method called Transport Layer Security Inspection (TLSI), which involves decrypting packets to do full packet inspection. TLSI is expensive in computational performance and complexity, and over and above all, it violates the users' privacy. Researchers from Cisco have proposed that it is possible to identify malicious encrypted traffic by techniques other than TLSI and that the unencrypted TLS handshake messages, certificates, and flow metadata of malicious traffic are distinct from benign. These differences can be effectively used in machine learning to classify malicious and benign encrypted traffic [35]. This dissertation aims to assess the feasibility and effectiveness of the proposed alternative to TLSI. We sourced thousands of malware and benign flows and then used the Cisco tool called Joy to extract the features from the unencrypted TLS handshake messages, certificates, and flow metadata. To understand the characteristic behaviour between malicious and benign flows, we did a data exploration, summarized the unique values of the features from our datasets, and compared them with the feature values from the Cisco datasets used in the research paper [35]. We then selected features that had the most differentiating power in our dataset. The selected features were inputs into the two supervised classifiers: logistic regression and random forest. The classifiers were trained and tested on the offline datasets of benign and malware features, and we observed that the random forest performed better with an average accuracy of 98.92%. We concluded that it is viable and effective to use alternative techniques to detect HTTPS malware without TLSI.
Open Access
DevelopinThe Bayesian Description Logic BALC
(2018) Botha, Leonard; Meyer, Thomas; Peñaloza, Rafael
Description Logics (DLs) that support uncertainty are not as well studied as their crisp alternatives. This limits their application in many real world domains, which often require reasoning about uncertain or contradictory information. In this thesis we present the Bayesian Description Logic BALC, which takes existing work on Bayesian Description Logics and applies it to the classical Description Logic ALC. We define five reasoning problems for BALC; two versions of concept satisfiability (called total and partial respectively), knowledge base consistency, three subsumption problems (positive subsumption, p-subsumption, exact subsumption), instance checking, and the most likely context problem. Consistency, satisfiability, and instance checking have not previously been studied in the context of contextual Bayesian DLs and as such this is new work. We then go on to provide algorithms that solve all of these reasoning problems, with the exception of the most likely context problem. We found that all reasoning problems in BALC are in the same complexity class as their classical variants, provided that the size of the Bayesian Network is included in the size of the knowledge base. That is, all reasoning problems mentioned above (excluding most likely context) are exponential in the size of the knowledge base and the size of the Bayesian Network.
Open Access
Investigating how South African humanities researchers engage with digital archives
(2021) Mtombeni, Khanyisa; Suleman, Hussein
OBJECTIVE: Despite technological developments in the Digital Humanities space, it is unclear that the facilities offered by digital archives support the needs of Humanities researchers in developing countries. The purpose of this thesis is to investigate how South African Humanities scholars use digital archives in their research as well as in teaching and other academic activities. METHODS: This thesis utilizes non-random convenience sampling. A feature determination study provided the sampling frame, defined the scope for the survey tool, and was used to uncover trends in digital archives development in South Africa. A self-administered online survey was conducted with Humanities researchers in South Africa to answer the research question. The thesis utilises basic descriptive statistics in its attempt to study and interpret the responses of participating researchers. RESULTS: 102 participants responded to the online survey. Despite many South African digital archives having the functionality to discover, browse and search collections, they are missing the features for collaboration, accessing and managing resources. Only 20% of the survey respondents are satisfied with South African digital archives' process of making content easy to find and accessible, whereas 48% of the respondents consider themselves users of complex digital resources, 44% have the knowledge and experience for using Digital Humanities tools and services, and more than 70% find technology to be useful for learning and teaching. CONCLUSIONS: The usage of archives and their functionalities vary widely. Users have stronger preferences for tools that support basic discovery and personal and collaborative research, but many consider existing support for basic features to be inadequate. In terms of advanced functionalities for managing digital resources, users are interested in these to varying levels, but the inadequate support means that these are still somewhat speculative.
Open Access
Investigating language preferences in improving multilingual Swahili information retrieval
(2022) Telemala, Joseph Philipo; Suleman, Hussein
Multilingual Information Retrieval (MLIR) systems are designed to retrieve information from multiple languages in response to a query posed in another language or in one of the languages in which a user is looking for information. Researchers have proposed several approaches for combining the results from individual result lists to produce a single result list. Some are heuristics, such as round-robin, in which a result is drawn from each result list one at a time until all lists are exhausted, while others are Machine Learning (ML)-based, in which a model is trained using a variety of features from the query and the required documents. These approaches strive for topical relevance, which is the most important goal in satisfying users' information needs. However, multilingual speakers exhibit a variety of behaviours, some of which are unique to certain individuals based on their historical, cultural, and linguistic backgrounds. Unfortunately, these behaviours are ignored in the current MLIR system design and implementation. Current MLIR systems present results that do not take people's language preferences into account when ranking results. Studies have shown that users have different language preferences based on their search topics – Topic-Language (T-L) preferences. This study proposes using T-L preferences to improve the relevance of the ranked MLIR results. To achieve this aim, we used a survey-based study to try to understand the information needs and Web search behaviour of Swahili-speaking Web users in Tanzania. One bold behaviour of such multilingual Web users that emerged is code-switching. Several factors, such as information context and search topic, were identified as reasons for such frequent language switching. We then created a prototype multilingual search engine with which users interacted in order to quantify how much the language of the query or the selected results is influenced by the search topic. We estimated the relationship between the topic of search and the language of the query and clicked results using the resulting query and click-through logs. The findings revealed that Swahili-speaking Web users have language preferences for certain topics. For example, Kiswahili was significantly preferred as a results language in only 9% of the examined topics, English was preferred in 26% of the topics, and there was no preference for language of results in the remaining 65% of the topics. Based on these findings, we created the T-L-based algorithm, which re-ranks the results based on T-L associations/preferences. We evaluated our proposed T-L-based algorithm using clickthrough logs from our prototype guided multilingual search engine. The results show that incorporating language preferences into the ranking model significantly improves the relevance MLIR results in some specific cases. The strength of the T-L association and the number of relevant results in the preferred language's list were discovered to be driving factors in the performance improvement of the T-L-based algorithm. This thesis provides evidence that using language preferences can potentially improve the relevance of MLIR results for some topics that are preferentially expressed in specific languages. This is important in communities where information search and access are hampered by a variety of factors and there is a clear lineage in language use, where MLIR's topical relevance alone may not be sufficient.
Open Access
Mobile health data: investigating the data used by an mHealth app using different mobile app architectures
(2018) Faker, Faizel; Suleman, Hussein; deRenzi Brian
Mobile Health (mHealth) has come a long way in the last forty years and is still rapidly evolving and presenting many opportunities. The advancements in mobile technology and wireless mobile communication technology contributed to the rapid evolution and development of mHealth. Consequently, this evolution has led to mHealth solutions that are now capable of generating large amounts of data that is synchronised and stored on remote cloud and central servers, ensuring that the data is distributable to healthcare providers and available for analysis and decision making. However, the amount of data used by mHealth apps can contribute significantly to the overall cost of implementing a new or upscaling an existing mHealth solution. The purpose of this research was to determine if the amount of data used by mHealth apps would differ significantly if they were to be implemented using different mobile app architectures. Three mHealth apps using different mobile app architectures were developed and evaluated. The first app was a native app, the second was a standard mobile Web app and the third was a mobile Web app that used Asynchronous JavaScript and XML (AJAX). Experiments using the same data inputs were conducted on the three mHealth apps. The primary objective of the experiments was to determine if there was a significant difference in the amount of data used by different versions of an mHealth app when implemented using different mobile app architectures. The experiment results demonstrated that native apps that are installed and executed on local mobile devices used the least amount of data and were more data efficient than mobile Web apps that executed on mobile Web browsers. It also demonstrated that mobile apps implemented using different mobile app architectures will demonstrate a significant difference in the amount of data used during normal mobile app usage.
Open Access
Neuro-evolution search methodologies for collective self-driving vehicles
(2019) Huang, Chien-Lun Allen; Nitschke, Geoff
Recently there has been an increasing amount of research into autonomous vehicles for real-world driving. Much progress has been made in the past decade with many automotive manufacturers demonstrating real-world prototypes. Current predictions indicate that roads designed exclusively for autonomous vehicles will be constructed and thus this thesis explores the use of methods to automatically produce controllers for autonomous vehicles that must navigate with each other on these roads. Neuro-Evolution, a method that combines evolutionary algorithms with neural networks, has shown to be effective in reinforcement-learning, multi-agent tasks such as maze navigation, biped locomotion, autonomous racing vehicles and fin-less rocket control. Hence, a neuro-evolution method is selected and investigated for the controller evolution of collective autonomous vehicles in homogeneous teams. The impact of objective and non-objective search (and a combination of both, a hybrid method) for controller evolution is comparatively evaluated for robustness on a range of driving tasks and collection sizes. Results indicate that the objective search was able to generalise the best on unseen task environments compared to all other methods and the hybrid approach was able to yield desired task performance on evolution far earlier than both approaches but was unable to generalise as effectively over new environments.
Open Access
Ontology verbalization in agglutinating Bantu languages: a study of Runyankore and its generalizability
(2019) Byamugisha, Joan; Keet, Catharina Maria; Brian DeRenzi
Natural Language Generation (NLG) systems have been developed to generate text in multiple domains, including personalized patient information. However, their application is limited in Africa because they generate text in English, yet indigenous languages are still predominantly spoken throughout the continent, especially in rural areas. The existing healthcare NLG systems cannot be reused for Bantu languages due to the complex grammatical structure, nor can the generated text be used in machine translation systems for Bantu languages because they are computationally under-resourced. This research aimed to verbalize ontologies in agglutinating Bantu languages. We had four research objectives: (1) noun pluralization and verb conjugation in Runyankore; (2) Runyankore verbalization patterns for the selected description logic constructors; (3) combining the pluralization, conjugation, and verbalization components to form a Runyankore grammar engine; and (4) generalizing the Runyankore and isiZulu approaches to ontology verbalization to other agglutinating Bantu languages. We used an approach that combines morphology with syntax and semantics to develop a noun pluralizer for Runyankore, and used Context-Free Grammars (CFGs) for verb conjugation. We developed verbalization algorithms for eight constructors in a description logic. We then combined these components into a grammar engine developed as a Protégé5X plugin. The investigation into generalizability used the bootstrap approach, and investigated bootstrapping for languages in the same language zone (intra-zone bootstrappability) and languages across language zones (inter-zone bootstrappability). We obtained verbalization patterns for Luganda and isiXhosa, in the same zones as Runyankore and isiZulu respectively, and chiShona, Kikuyu, and Kinyarwanda from different zones, and used the bootstrap metric that we developed to identify the most efficient source—target bootstrap pair. By regrouping Meinhof’s noun class system we were able to eliminate non-determinism during computation, and this led to the development of a generic noun pluralizer. We also showed that CFGs can conjugate verbs in the five additional languages. Finally, we proposed the architecture for an API that could be used to generate text in agglutinating Bantu languages. Our research provides a method for surface realization for an under-resourced and grammatically complex family of languages, Bantu languages. We leave the development of a complete NLG system based on the Runyankore grammar engine and of the API as areas for future work.
Open Access
Predicting household poverty with machine learning methods: the case of Malawi
(2022) Chinyama, Francis; Berman, Sonia
Poverty alleviation continues to be paramount for developing countries. This necessitates the need for poverty tracking tools to monitor progress towards this goal and effect timely interventions. One major way poverty has been tracked in Malawi is by carrying out integrated household surveys every five years to quantify poverty at local and national levels. However, such surveys have been documented as very expensive, tedious, and sparsely administered by many low- and middle-income countries. Therefore, this study looked at whether machinelearning models can be used on existing survey data to predict poor and non-poor households, and whether these models can predict poverty using a smaller number of features than those collected in integrated household surveys. This was achieved by comparing the performance of three off-the-shelf, open-source machinelearning classification algorithms namely Logistic Regression, Extra Gradient Boosting Machine and Light Gradient Boosting Machine, in correctly predicting poor and non-poor households from Malawi survey data. These supervised learning algorithms were trained using 10-fold cross-validation. The experiments were carried out on the full panel of features which represent all the questions asked in a household survey, as well as on smaller feature subsets. The Filter method and SHapley Additive exPlanations method were used to rank the importance of the features, and smaller data subsets were selected based on these rankings. The highest prediction accuracy achieved for the full panel data set of 486 features was 87%. When the Filter method rankings were used, the models' prediction accuracy dropped to 63% for the top 50 features subset. However, using the SHAP method rankings, the maximum prediction accuracy level was maintained and only dropped slightly to 86% with the top 50 feature subset; to 84% with the top 20 features; and 73% for the top 10 features. Area under the Curve, Receiver Operating Characteristic curve, recall, precision, F1 score, Matthews Correlation Coefficient and Cohen's Kappa scores confirmed the classification models' reliability. The study, therefore, established that poverty can be predicted by open-source machine learning algorithms using a substantially reduced number of features with accuracy comparable to using the full feature set. A policy recommendation is to employ only the top explanatory features in surveys. This will enable shorter, lower-cost surveys that can be administered more frequently. The aim is to assist policymakers and aid organisations to make more timely interventions with better targeting of the poorest.
Open Access
Robust and cheating-resilient power auctioning on Resource Constrained Smart Micro-Grids
(2018) Marufu, Mufudzi Anesu Chapman; Meyer, Thomas
The principle of Continuous Double Auctioning (CDA) is known to provide an efficient way of matching supply and demand among distributed selfish participants with limited information. However, the literature indicates that the classic CDA algorithms developed for grid-like applications are centralised and insensitive to the processing resources capacity, which poses a hindrance for their application on resource constrained, smart micro-grids (RCSMG). A RCSMG loosely describes a micro-grid with distributed generators and demand controlled by selfish participants with limited information, power storage capacity and low literacy, communicate over an unreliable infrastructure burdened by limited bandwidth and low computational power of devices. In this thesis, we design and evaluate a CDA algorithm for power allocation in a RCSMG. Specifically, we offer the following contributions towards power auctioning on RCSMGs. First, we extend the original CDA scheme to enable decentralised auctioning. We do this by integrating a token-based, mutual-exclusion (MUTEX) distributive primitive, that ensures the CDA operates at a reasonably efficient time and message complexity of O(N) and O(logN) respectively, per critical section invocation (auction market execution). Our CDA algorithm scales better and avoids the single point of failure problem associated with centralised CDAs (which could be used to adversarially provoke a break-down of the grid marketing mechanism). In addition, the decentralised approach in our algorithm can help eliminate privacy and security concerns associated with centralised CDAs. Second, to handle CDA performance issues due to malfunctioning devices on an unreliable network (such as a lossy network), we extend our proposed CDA scheme to ensure robustness to failure. Using node redundancy, we modify the MUTEX protocol supporting our CDA algorithm to handle fail-stop and some Byzantine type faults of sites. This yields a time complexity of O(N), where N is number of cluster-head nodes; and message complexity of O((logN)+W) time, where W is the number of check-pointing messages. These results indicate that it is possible to add fault tolerance to a decentralised CDA, which guarantees continued participation in the auction while retaining reasonable performance overheads. In addition, we propose a decentralised consumption scheduling scheme that complements the auctioning scheme in guaranteeing successful power allocation within the RCSMG. Third, since grid participants are self-interested we must consider the issue of power theft that is provoked when participants cheat. We propose threat models centred on cheating attacks aimed at foiling the extended CDA scheme. More specifically, we focus on the Victim Strategy Downgrade; Collusion by Dynamic Strategy Change, Profiling with Market Prediction; and Strategy Manipulation cheating attacks, which are carried out by internal adversaries (auction participants). Internal adversaries are participants who want to get more benefits but have no interest in provoking a breakdown of the grid. However, their behaviour is dangerous because it could result in a breakdown of the grid. Fourth, to mitigate these cheating attacks, we propose an exception handling (EH) scheme, where sentinel agents use allocative efficiency and message overheads to detect and mitigate cheating forms. Sentinel agents are tasked to monitor trading agents to detect cheating and reprimand the misbehaving participant. Overall, message complexity expected in light demand is O(nLogN). The detection and resolution algorithm is expected to run in linear time complexity O(M). Overall, the main aim of our study is achieved by designing a resilient and cheating-free CDA algorithm that is scalable and performs well on resource constrained micro-grids. With the growing popularity of the CDA and its resource allocation applications, specifically to low resourced micro-grids, this thesis highlights further avenues for future research. First, we intend to extend the decentralised CDA algorithm to allow for participants’ mobile phones to connect (reconnect) at different shared smart meters. Such mobility should guarantee the desired CDA properties, the reliability and adequate security. Secondly, we seek to develop a simulation of the decentralised CDA based on the formal proofs presented in this thesis. Such a simulation platform can be used for future studies that involve decentralised CDAs. Third, we seek to find an optimal and efficient way in which the decentralised CDA and the scheduling algorithm can be integrated and deployed in a low resourced, smart micro-grid. Such an integration is important for system developers interested in exploiting the benefits of the two schemes while maintaining system efficiency. Forth, we aim to improve on the cheating detection and mitigation mechanism by developing an intrusion tolerance protocol. Such a scheme will allow continued auctioning in the presence of cheating attacks while incurring low performance overheads for applicability in a RCSMG.
Open Access
The effects of augmented virtuality on presence, workload, and input performance
(2018) Clarkson, Jacob; Blake, Edwin
Head-Mounted Displays (HMDs) offer, more than any easily accessible technology that has come before, the sensation of presence – that feeling that you are “really there” in a virtual world. However, HMDs cut the wearer off from the real world, making even trivial interactions, such as having a drink or typing, difficult and frustrating. In the home context where these devices are most likely to be used, such interactions are commonplace, and in order to execute them, users have to remove the HMD (“peep”), breaking their sense of presence. How, then, can real-world interactions during HMD usage be facilitated such that presence is damaged as little as possible? Previous work indicates that Augmented Virtuality (AV), a technique that allows the wearer of an HMD to see through it when they need to, is a promising answer to this question. However, direct comparisons between AV and VR that thoroughly account for presence and workload are lacking. To corroborate previous findings, and to address some of the gaps in the current literature, we conducted a quantitative user experiment to compare our own implementation of AV to VR in terms of presence, workload, and typing performance. The experiment followed a betweengroups design with participants selected via pseudo-random convenience sampling of university students. To simulate the context of home usage – an extended immersive session that must occasionally be interrupted – we designed a mixed reality game that periodically required the player to interact with real-world objects before they could proceed. Participants in the experimental group played the game using our AV system to assist them in completing the required real-world tasks. Participants in the control group used pure VR to play the game and had to peep. This allowed us to directly compare AV to VR in terms of the levels of presence and workload experienced. These data were gathered using post-hoc self-report questionnaires. To measure and compare typing performance under various conditions, we created desktop, VR, and AV versions of a typing test that participants had to complete. We found that typing performance in AV was significantly better than in VR, but did not reach the levels achieved in baseline desktop conditions. While there was not a significant difference in the overall level of workload associated with using AV compared to VR, participants in the AV condition were able to interact successfully with the real world without having to remove the HMD, and reported being significantly less frustrated than those in the VR condition. Finally, AV users reported significantly higher levels of presence than those who used VR.
Open Access
The performance of coevolutionary topologies in developing competitive tree manipulation strategies for symbolic regression
(2020) Ombura, Martin; Nitschke, Geoff Stuart
Computer bugs and tests are antagonistic elements of the software development process, with the former attempting to corrupt a program and the latter aiming to identify and fix the introduced faults. The automation of bug identification and repair schemes through automated software testing is an area of research that has only seen success in niche areas of software development but has failed to progress into general areas of computing due to the complexity and diversity of programming languages, codebases and developer coding practices. Unlike traditional engineering fields such as mechanical or civil where project specifications are carefully outlined and built towards, software engineering suffers from a lack of global standardization required to “build from a spec”. In this study we investigate a coevolutionary spec-based approach to dynamically damage and repair programs mathematical programs (functions). We opt for mathematical functions instead of software due to their functional similarities and simpler syntax and semantics. We utilize symbolic regression (SR) as a framework to analyze the error maximized by bugs and minimized by test. We adopt a hybrid evolutionary algorithm (EA) that implements the tree based phenotypic structure of genetic programming (GP) and the list-based chromosome of genetic algorithm (GA) that permits embedding of mathematical tree manipulation (MTM) strategies, as well as adequate selection mechanisms for search. Bugs utilize the MTM strategies in their chromosome to manipulate the input program (IP) with the aim of maximizing the error while tests adopt a set of their own MTM strategies to repair the damaged program using a spec generated from the IP to guide the repair process. Both adversarial agents are investigated in four common coevolutionary topologies, Hall of Fame (HoF), K-Random Tournaments (KRT), Round Robin (RR) and Single Elimination Tournament (SET). We ran 1556 simulations each generating a random polynomial that the bugs and tests would have to contend over in all 4 topologies. We observed that KRT with a low k value of 5 performs best from a computational and fitness standpoint for all bugs and tests. Bugs were dominant in nearly all topologies for all polynomial complexities, whereas tests struggled in the HoF, RR and SET topologies as the input programs became more complex. The competitive landscape however was quite chaotic with the best individuals lasting a maximum of 14 generations out of 300, with the average top individuals lasting only 1 generation. This made predictions on when the best individuals would be born nearly impossible as the coevolutionary landscape changed quite rapidly and non-deterministically. The kinds of MTM strategies selected by both bugs and tests depended on the level of complexity of the input programs. For input programs that had negative polynomials, the best bugs opted to delete the program entirely and build a completely new tree, whereas the best tests were unable to select viable specialized strategies to repair such programs. For programs that had large polynomial degrees, bugs opted for strategies that added nodes their underlying GP tree, in the hopes of damaging the input program more. Tests on the other hand implemented strategies to carefully reduce the complexity of the polynomial. Tests however, frequently overcompensated when attempting to fix the fit bugs, leading to mediocre solutions.
Open Access
The uncomfortable chair of the colonial past and racist present an effective approach to white discomfort in the Netherlands
(2022) Kolman, Kiki; Msomi, Zuziwe Nokwanda
Activists, artists, academics, and other experts have pointed out the problem of (manifestations of) white discomfort in the Netherlands: white Dutch people experience strong unease when topics like racism or the colonial past are addressed, resulting in defensiveness and avoidance of important conversations – hence, obstructing the anti-racism struggle. In order to tackle this problem, it is important to understand where white discomfort is rooted and how it manifests. Gloria Wekker's work shows the importance of ‘white innocence' in understanding Dutch whiteness – the false conception of (progressive) Dutch white people that the Netherlands and they themselves are innocent with regard to racism. This myth of innocence is constructed on a collective national level. International literature on white discomfort, however, is often (but not always) focused on the individual's psychology. Therefore, this interdisciplinary thesis explores an additional conception of white discomfort, grounded in the Dutch reality, that acknowledges the historical and collective context of whiteness. This is done by combining critical whiteness studies with Affect Theory, specifically Sara Ahmed's work on whiteness and discomfort. The result is an understanding of white discomfort as the friction between the historically shaped self-perceived innocent body, and the historically shaped and continuously changing space this body is in. The conceptualisation is then further developed in a dance between theory and practice by analysing the results of focus groups with (self-identified) progressive white Dutch people. The research points to white discomfort as an interplay of 1) the identification of the individual with the collective of white Dutch people, or simply with the Netherlands as a country; 2) the collision with the innocent self-image when this collective collides with the innocent self-image; 3) the white person's desire to either be good or be perceived as good (most likely, a combination of both); And 4) the restricted space to speak in conversations on racism. The approach of white discomfort as affect, then, offers the opportunity to connect the white individual to the collective history of the Netherlands, while still acknowledging individual responsibility. And it helps uncover how the avoidance of white discomfort by individuals maintain problematic forms of whiteness.