https://www.um.edu.mt/library/oar/handle/123456789/8422 2026-04-06T06:58:21Z 2026-04-06T06:58:21Z https://www.um.edu.mt/library/oar/handle/123456789/95971 2022-05-19T08:08:20Z 2015-01-01T00:00:00Z

Title: Using social media as a basis for marketing initiatives Abstract: In computing, social media is represented as a mathematical graph which depicts users as nodes, and interactions as edges. Users are either persons or company profiles - even a specific product. The edges convey feelings, likings and acceptance which are known to be very useful for businesses to understand better their product's perception by users and consequently influence the business's marketing tactics. Due to the remarkable popularity reached by social media platforms, the data available is so vast that many marketing experts cannot ascribe significant meaning when looking at it as it is. The purpose of this project is to tap into this data from two popular social media platforms, and transform and process it in a way to enable deeper knowledge extraction that is usable in a marketing initiative. The final artefact of this project acquires data from the two platforms by connecting, and sending requests to their APIs. These requests return responses consisting of data available with regards to that particular request. This data is then processed to create one consolidated model encompassing the data acquired from both sources. This is then stored in a graph based data model which is queried as required. The data returned by the APIs highly depends on the social media platform itself as they have included further privacy concerns in their latest APIs which limits the data in the response. This artefact works with the given public data, however if users give permission to use their data in the application, it would be able to give more insightful information. The end result of this artefact consists of a number of charts which portray significant information that help business users to understand their fans and hence improve their marketing campaign. Description: B.Sc. IT (Hons)(Melit.)

2015-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/95969 2022-05-19T08:04:01Z 2015-01-01T00:00:00Z

Title: Investigating the recognition of the static hand signs used by fingerspelling on smart phone Abstract: One of the leading challenges faced by the deaf members of society is the communication barrier which arises due to their lack of hearing and the lack of individuals who are able to express themselves through sign language. When considering both the deaf and the hearing communities, one factor common to both parties is the use of smart phone devices. However, whereas hearing individuals are able to freely communicate through audio calls on a smart phone, the latter is not the case for deaf individuals who rely on sign language as their principal means of communication. This thesis attempts to tackle this issue by correlating smart phone technologies with a subset of sign language communication. Specifically, it attempts to develop an automated system which recognises the various static hand postures which may be found within a sign language's alphabet through the use of a smart phone device. This is done through an investigation of the existing technologies in the field of study, identification of possible challenges which may come up, as well as possible solutions on how to tackle them and finally the development of a smart phone application with the knowledge gained through this research. The main lifecycle for the recognition of these static hand postures involves passing an image through a series of image processing and feature extraction techniques, followed by a machine learning method that is used to recognize the static symbols captured by the camera. The recognition rate of the developed proof of concept is tested, which tops off at 83.7% average recognition. Through the tests performed, it was also concluded that this value may be further increased through the use of a larger training dataset and different extraction algorithms. Description: B.Sc. IT (Hons)(Melit.)

2015-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/95186 2022-05-06T08:08:56Z 2015-01-01T00:00:00Z

Title: Comparing Erlang-based web technologies with emerging MEAN stack for scalable web applications Abstract: Online services such as social networks, e-commerce sites and multiplayer games experience heavy user traffic on a daily basis and users expect instant response times irrespective of the number of concurrent users being served. To meet such expectations, web applications should be able to scale and perform well under heavy loads. The selection of underlying technologies is of paramount importance, based on a good understanding of expected usage scenarios and transaction types. This study acknowledges the fact that the emerging JavaScript-based MEAN stack (MongoDB, Express, AngularJS, Node.js) is portrayed as the natural choice for building scalable web applications. Despite this, Erlang - a programming language developed in 1986 - was specifically designed to handle massive concurrency and has been shown to withstand heavy loads in the telecommunications domain. This study proposes an Erlang-based alternative to the MEAN stack. The resulting stack is compared with the MEAN stack through a series of lab-based experiments to quantify the benefits of using one stack over another in specific scenarios. Tests assess different types of transactions, including (1) HTTP requests for static content, (2) HTTP requests with database reads/writes and (3) bi-directional communication via the WebSocket protocols. Different measurements are considered in the evaluation process, including resource consumption, response times and the actual throughput handled. The results obtained provide indications of how each stack scales and performs in specific scenarios with respect to the different measures being considered. Description: B.Sc. IT (Hons)(Melit.)

2015-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/95010 2022-05-05T07:14:41Z 2015-01-01T00:00:00Z

Title: Spam detection using machine learning techniques Abstract: The introduction of electronic mail brought about a reliable and economical method of communication. However, apart from the advantages it offered, several disadvantages also came along, one of which is spam. Spam is unsolicited mail that, in some way or another, finds its way into our inboxes. Throughout the years, spam has increased and developed ways of disguising itself as a legitimate e-mail through deceptive appearances. More significantly, the last few decades have seen a vast production of techniques, designated to recognize and block spam from reaching our inboxes. A set of rules and Machine Learning Algorithms have been tried and tested as anti-spam filters to reduce as much spam as possible. This dissertation explores the comparison between a developed set of artefacts, together with the Machine Learning algorithms that have been tested in El-Sayed ElAlfy' s paper Learning Methods for Spam Filtering in the collection of research papers found in Computer Systems, Support and Technology (2011). Using the Spambase dataset to train and test the algorithms, 2 artefacts were created with the potential of being Anti-spam filters. Artefact 1 proposes a method that composed ofN-grams and the use of the entropy technique, combined with the Naive Bayes algorithm. Artefact 2 is a notable neural-network algorithm which is known as the Error Backpropagation Neural Network. Artefact 1 and 2 were compared with other well-known Machine Learning algorithms such as Support Vector Machine, Multi-Layer Perceptron, k-NN, decision trees and others. With regards to results, Artefact 1 ranked first in precision at 96.16% when compared with the other 20 algorithms tested, followed by the Radial Basis Function algorithm at 95.52%. Artefact 2, with a precision of 93.44% was ranked 18th, followed by the Nai"ve Bayes algorithm with a precision of 93.29%. The accuracy achieved by Artefact 1 was that of 89.66% and for Artefact 2 this was 92.15%. From the aforementioned results, it can be seen that Artefact 1 can offer quite competitive results. Description: B.Sc. IT (Hons)(Melit.)

2015-01-01T00:00:00Z