Software Engineering Teaching Unit
Permanent URI for this collectionhttp://repository.kln.ac.lk/handle/123456789/26470
Browse
6 results
Search Results
Item Change detection in dynamic attributed networks(2018) Hewapathirana, I.U.A network provides powerful means of representing complex relationships between entities by abstracting entities as vertices, and relationships as edges connecting vertices in a graph. Beyond the presence or absence of relationships, a network may contain additional information that can be attributed to the entities and their relationships. Attaching these additional attribute data to the corresponding vertices and edges yields an attributed graph. Moreover, in the majority of real-world applications, such as online social networks, financial networks and transactional networks, relationships between entities evolve over time. Change detection in dynamic attributed networks is an important problem in many areas, such as fraud detection, cyber intrusion detection, and health care monitoring. It is a challenging problem because it involves a time sequence of attributed graphs, each of which is usually very large and can contain many attributes attached to the vertices and edges, resulting in a complex, high-dimensional mathematical object. In this survey we provide an overview of some of the existing change detection methods that utilize attribute information. We categorize these methods based on the levels of structure in the graph that are exploited to detect changes. These levels are vertices, edges, subgraphs, communities, and the overall graph. We focus attention on the strengths and weaknesses of these methods, including their performance and scalability. Furthermore, we discuss some publicly available dynamic network datasets and give a brief overview of models to generate dynamic attributed networks. Finally, we discuss the limitations of existing approaches identifying key areas for future research.Item Change detection in noisy dynamic networks: a spectral embedding approach(2020) Hewapathirana, I.U.; Lee, Dominic; Moltchanova, Elena; McLeod, JeanetteChange detection in dynamic networks is an important problem in many areas, such as fraud detection, cyber intrusion detection and healthcare monitoring. It is a challenging problem because it involves a time sequence of graphs, each of which is usually very large and sparse with heterogeneous vertex degrees, resulting in a complex, high-dimensional mathematical object. Spectral embedding methods provide an effective way to transform a graph to a lower dimensional latent Euclidean space that preserves the underlying structure of the network. Although change detection methods that use spectral embedding are available, they do not address sparsity and degree heterogeneity that usually occur in noisy real-world graphs and a majority of these methods focus on changes in the behaviour of the overall network. In this paper, we adapt previously developed techniques in spectral graph theory and propose a novel concept of applying Procrustes techniques to embedded points for vertices in a graph to detect changes in entity behaviour. Our spectral embedding approach not only addresses sparsity and degree heterogeneity issues, but also obtains an estimate of the appropriate embedding dimension. We call this method CDP (change detection using Procrustes analysis). We demonstrate the performance of CDP through extensive simulation experiments and a real-world application. CDP successfully detects various types of vertex-based changes including (1) changes in vertex degree, (2) changes in community membership of vertices, and (3) unusual increase or decrease in edge weights between vertices. The change detection performance of CDP is compared with two other baseline methods that employ alternative spectral embedding approaches. In both cases, CDP generally shows superior performance.Item Combining Information from Multiple Views for Vertex-Based Change Detection in Dynamic Networks: A Comparative Study(2022) Hewapathirana, I.U.In the majority of previous network-based change detection methods, entities are often represented as vertices in an undirected graph with weighted edges describing the relationships between entities. However, in most real-world applications, entities share different types of relationships forming multi-view networks that can be well represented by multiple undirected graphs over the same set of vertices. Integrating information across multiple undirected graphs for change detection in noisy dynamic networks is a crucial but challenging issue. In this paper, a multi-view dynamic network is represented as a time sequence of tensors by encoding multi-graph information into the slices of the tensor. We employ spectral embedding together with Procrustes analysis to detect changes due to vertex behavior, that is, vertex-based changes. Through extensive simulation experiments, we demonstrate the performance of several strategies to combine the information from different slices of the tensor, and obtain a single embedding of the multi-view network. In all experiments, we compare the performance of these methods over a variety of changes ranging from easy to difficult for the purpose of detecting vertices that undergo change. We show that two strategies, MCDP-I and MCDP-II, successfully detect all types of vertex-based changes that were considered in the experiments and show better performance compared to the others when the connectivity structure varied considerably across the slices of the tensor. MCDP-I employs a higher-order singular value decomposition to factorize the tensor at each time instant, and MCDP-II, initially embeds each slice of the tensor separately using matrix factorization and then applies generalized Procrustes analysis techniques on the resulting set of embeddings to obtain a combined embedding. Finally, we also illustrate the performance of MCDP-I and MCDP-II for a real-data application.Item Utilizing Prediction Intervals for Unsupervised Detection of Fraudulent Transactions: A Case Study(2022) Hewapathirana, I.U.Money laundering operations have a high negative impact on the growth of a country’s national economy. As all financial sectors are increasingly being integrated, it is vital to implement effective technological measures to address these fraudulent operations. Machine learning methods are widely used to classify an incoming transaction as fraudulent or non-fraudulent by analyzing the behaviour of past transactions. Unsupervised machine learning methods do not require label information on past transactions, and a classification is made solely based on the distribution of the transaction. This research presents three unsupervised classification methods: ordinary least squares regression-based (OLS) fraud detection, random forest-based (RF) fraud detection and dropout neural network-based (DNN) fraud detection. For each method, the goal is to classify an incoming transaction amount as fraudulent or non-fraudulent. The novelty in the proposed approach is the application of prediction interval calculation for automatically validating incoming transactions. The three methods are applied to a real-world dataset of credit card transactions. The fraud labels available for the dataset are removed during the model training phase but are later used to evaluate the performance of the final predictions. The performance of the proposed methods is further compared with two other unsupervised state-of-the-art methods. Based on the experimental results, the OLS and RF methods show the best performance in predicting the correct label of a transaction, while the DNN method is the most robust method for detecting fraudulent transactions. This novel concept of calculating prediction intervals for validating an incoming transaction introduces a new direction for unsupervised fraud detection. Since fraud labels on past transactions are not required for training, the proposed methods can be applied in an online setting to different areas, such as detecting money laundering activities, telecommunication fraud and intrusion detection.Item A Review on Current Trends and Applications of Social Media Research in Sri Lanka(2023) Hewapathirana, I.U.Standard research on social media and its applications has been widely disseminated in developed nations. But in Sri Lanka, research in this area has been released far less frequently. However, social media usage in the country is evolving regardless of age, sex, education level, or other limitations. This study aims to fill the gap by conducting a comprehensive review of social media-based research conducted in Sri Lanka between 2012 and 2022. A systematic search of reputable databases, including IEEE Xplore, ScienceDirect, Emerald Insight, Google Scholar, and Springer Link, identified 57 relevant papers for analysis. The review highlights the diversity of application areas where social media research has been employed in Sri Lanka, including disaster management, public health, marketing, education, and more. Additionally, the analysis highlights the methodological approaches employed in social media analytics and the specific social media platforms utilized by researchers in Sri Lanka. The results of the current study serve as a timely resource, enabling policymakers and decision-makers to identify the potential avenues of social media research in Sri Lanka. By understanding the existing trends and implications, stakeholders can harness the power of social media data to make informed policy decisions, develop effective marketing strategies, enhance public health initiatives, and revolutionize educational practices.Item Analysis of Employee Satisfaction using Artificial Neural Networks: A Case Study in the Information Technology Industry in Sri Lanka(2023) Hewapathirana, I.U.; Thilakarathna, W.A.S.M.S.Job satisfaction is vital to the prosperity of all industries, including the information technology (IT) sector. This research represents the pioneering attempt to employ Artificial Neural Networks (ANNs) for the purpose of examining the factors that affect job satisfaction among IT professionals in Sri Lanka. Data was gathered from a survey of 156 IT professionals and was analyzed statistically to identify seven factors that influence job satisfaction. An ANN was trained to predict job satisfaction using the extracted factors as inputs. The accuracy of the model in predicting job satisfaction was 94.64%. This suggests that ANNs have the potential to identify critical factors for IT employees and target interventions to increase their satisfaction. Additional research on the fitted ANN revealed that working conditions, family-friendly policies, and level of autonomy at work are crucial factors in determining a person's job satisfaction. Employers can use our findings to increase employee satisfaction by implementing appropriate policies. In addition, our ANN model can used to identify employees who are likely to leave and to provide them with customized interventions.