/library/oar/handle/123456789/106874 2025-11-09T07:10:28Z /library/oar/handle/123456789/122825 Title: SADIP : semi-automated data integration system for protein databases Abstract: Biologists must commonly combine information from different biological databases, by manually following cross-references (hyperlinks), using the distinct access methods and data formats provided by the databases. Past research in data integration has outlined several approaches which can integrate biological databases to provide a unified view. One approach is known as data warehousing. The current state of the art in biological data warehousing, requires bespoke software development and maintenance for each database. In our view, this is infeasible given the large number of constantly changing biological databases with varying access methods and data formats. This project aims to develop a tool which can automatically integrate biological information from different databases into a data warehouse, using user-defined configurations. This tool was applied to construct a property graph database with integrated information from 10 protein databases. This allows bioinformaticians to specify complex queries through the Standard Query Language (SQL). On top of this, a web-based user interface was developed which provides biologists with all integrated information related to a single protein identified by a UniProtKB identifier. The obtained results for the utilised configuration show that developing such a tool is feasible. However, the developed prototype requires further amendments to improve its flexibility, robustness, and security. Further results obtained show that the data warehouse provides biologists with a considerable amount of valuable information but should be extended to incorporate a wider variety of biological information. Finally, the results highlighted performance deficiencies for nested information and structural domains. Description: B.Sc. IT (Hons)(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/107190 Title: Assessing the feasibility of tokenisation in healthcare Abstract: External organisations, hospitals, public health authorities, general practitioners, and insurance companies often want to access patients’ health data for research purposes, treatments, service improvement. Also, because the patient went to this new hospital and this new hospital needs to access their data currently, in order for the new doctor to access your data from another hospital, a consent form must be signed by the patient, indicating what data will be shared, with whom, and for how long. This dissertation proposes a framework to tokenise patients’ health data. It will facilitate external organisations to access patients’ health data on a decentralised database only after authorisation. Therefore, the patient can share their data with the external organisation after receiving tokens. Patients may use tokens to pay for the hospital’s service. The patient could either buy tokens, they could have tokens from a previous visit, or they may have received tokens for sharing their data with external organisations and will pay the hospital with those tokens. Besides tokenising patients’ health data and giving them an incentive to share their data, they also gain complete control over it. With the use of smart contracts, the patient will be able to decide with whom their data will be shared and will also be able to specify which health data the external entity may gain access to. This will allow patients to monitor who is accessing their data. This dissertation developed a custom smart contract for healthcare professionals to enter the patient's data. The smart contract is set up so that their data has changed every time the patients go to the hospital. A new asset will be created to separate the different visits and allow patients to view their data more clearly. A survey was carried out to find out the general public’s opinion on healthcare data and the different uptakes of tokens this survey collected and analysed 151 responses. A related survey was collected and analysed 46 healthcare professionals’ responses to determine what they would do if the health data were tokenised and how they need health data to work. The data stored on the blockchain will be more secure, and smart contracts will facilitate efficient data sharing. Smart contracts will also allow the patient to specify what data may be shared and have a list of all the external organisations with which their data was shared. The smart contracts were implemented in this dissertation using the Hyperledger Fabric blockchain and have a couple of functions that may be used to create, update and delete assets and view these assets. These functions were tested to show the different types of errors that could be displayed and to show what can and cannot be entered when using the functions. Then an analysis of the survey questions was done, explaining what some of the charts meant to this dissertation and also showing how useful this dissertation could be in the real world. One of the charts from the general public survey shows that in fact, a quarter of the people answering the survey did not even want to share their data with anyone. Another chart showed that 73.7% would like to know who is accessing their data, and because of this result detail in the smart contract makes a list of all the external organisations accessing their data. This dissertation tried to develop a framework to tokenise patients’ health data so that patients’ will gain full control of their health data and get an incentive for sharing their data. This was done with smart contracts, Hyperledger Fabric, to be able to put patients’ data on a decentralised database so that their data is more secure, their data will be interchangeable, will have more integrity and interoperability and will have real-time updating and accessing. A BPMN diagram was created to help the reader understand all the scenarios in which tokenisation may be used and help improve healthcare. Primary research was gathered and analysed, in order to see the different ideas of healthcare professionals on the uptake of tokens and to better understand what the general public would want to do with their data. A smart contract was implemented and tested and has different functions that can be used to control the data going in and out of the system. Description: B.Sc. IT (Hons)(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/107188 Title: Analysing diverse algorithms performing music genre recognition Abstract: Nowadays, in order to discover music which matches our tastes we tend to rely greatly on the same applications we use to listen to the actual music, and in the process, allow algorithms to introduce us to new music genres that may be of interest to us. Because of this, it is especially important that these applications are able to achieve a good understanding of our listening habits, genre taste and connections with the performers, among other factors. A musical piece has various characteristics by which it can be described, and hence, songs with similar characteristics can be organised together in a single class, referred to as a musical genre. The challenge here is that the definition of a musical genre is itself subjective and the boundaries between one musical genre and another are not regularised and are rather based on user perception. This project seeks to determine the extent to which Music Genre Recognition can be performed by evaluating different Machine Learning algorithms used in the industry. The algorithms were applied to a curated benchmark of audio tracks with corresponding genre labels and compared to similar models documented in the literature. Therefore, to tackle this task, algorithms such as Artificial Neural Networks, Convolutional Neural Networks, Recurrent Neural Networks and Gradient Boosting Machines were created, experimented with and compared. Audio tracks were used as the dataset content of choice, to simulate as much as possible real-world applications, from which 13 Mel Frequency Cepstral Coefficients were extracted to be used as inputs for the algorithms, as these coefficients were found to be among the best features to approximate the human auditory system. The results obtained show that the Convolutional Neural Network and the relatively new Gradient Boosting Machine, namely XGBoost both have the best performance among the others. It was also discovered that small input samples of features are not only capable of training a classifying algorithm, but actually provide the best results. Description: B.Sc. IT (Hons)(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/107186 Title: Forecasting hospital resource requirements using remote-sensing satellite imagery Abstract: In recent years, the volume of satellite imagery has also grown significantly, as more satellites were placed in orbit to monitor the status of the Earth. Thus, the availability of this expanding satellite data presents an excellent opportunity for discovering novel approaches, specifically data-driven methodologies. Indeed, the Sentinel-3 Earth observation satellite developed by the European Space Agency as part of the Copernicus Programme was used to capture the land surface temperature data for this project in contrast to the ground data offered by the MET Office, Malta International Airport. Extracting land surface temperature for a particular region using satellite images is quite challenging, which the research project aims to cover. Hospital admissions and patient length of stay are highly dynamic. Different departments, particularly the emergency ward, may experience long and perpetual queues during specific periods. In such cases, the overall workflow for effectively managing and preparing medical departments’ resources has become a daunting task. Not only that, but economists also bear serious concern about the efficient use of scarce resources in hospitals. In certain countries, this issue is also reaching unsustainable levels endured by the general population [1-2]. Therefore, a growing need to forecast hospital resource requirements has become more apparent in recent years. Through sustainable use of modelling, the allocation of resources can be significantly improved so that the number of bed crises could be reduced to a minimum, or even avoided completely [3]. This research study attempts to solve such issues by closely identifying a tentative relationship between fluctuations in land surface temperature in conjunction with hospital discharge data offered by the Directorate for Health ¸£ÀûÔÚÏßÃâ·Ñ & Research (DHIR). The data covers anonymous information including the admission and discharge data and the admitting and discharging ward of every patient for a specific period. Land Surface Temperature (LST) data were obtained daily, spanning one year. Moreover, an inset figure was drawn on each LST data frame to cover the geographical region surrounding the Maltese borders solely. This was done to obtain the best possible accuracy when comparing the two datasets simultaneously. Moreover, these satellite images were compared and temperature changes were recorded from time to time. The results indicate an inverse relationship between hospital admissions and temperature increase, which could be helpful for health professionals and policymakers that may seek the impact of movements in temperature concerning hospital admission figures. The satellite data also indicated a positive relationship with ground temperature. It may also be beneficial for further research into the topic in question. Description: B.Sc. IT (Hons)(Melit.) 2022-01-01T00:00:00Z