OBJECTIVE: We aim to improve Drug-Drug Interactions (DDIs) by explicitly injecting medicinal-chemistry knowledge of functional groups (FGs) into graph neural network (GNN) message passing, in both transductive and inductive settings. Our goal is to (i) encode FG priors in a trainable way that enhances representation quality without handcrafting features, and (ii) yield interpretable attributions that align learned weights with pharmacologically meaningful FG patterns. METHODS: We introduce FG-DDI, a dual-view GNN that augments both intra- and inter-molecular reasoning. At the intra-molecular level, atom/bond messages are scaled by FG enrichment weights derived from detected FG motifs within each drug graph. At the inter-molecular level, a bipartite message-passing layer between a drug pair is modulated by FG-FG enrichment scores that reflect empirical co-occurrence in known DDIs. Enrichment is computed as odds ratios from corpus statistics and injected via learnable gates, ensuring differentiability and allowing data to override noisy priors. We couple this with standard supervision on interaction labels and report accuracy (ACC), AUROC, average precision (AP), and F1. Experiments use DrugBank (1706 drugs; 86 interaction types) and TwoSides (filtered triplets) under transductive and inductive splits (one unseen; both unseen). We perform ablations removing each FG term to isolate contributions and assess stability across splits. RESULTS: Comprehensive experiments on DrugBank and TwoSides datasets demonstrate that FG-DDI achieves superior performance compared to state-of-the-art methods. For DrugBank, the accuracy improves by 0.36% in transductive settings and by 0.46% and 1.42% in inductive settings, respectively for S1 and S2 partitioning. CONCLUSION: By systematically integrating chemical domain knowledge into deep learning architectures, this approach enables better generalization to unseen drug combinations while maintaining computational efficiency, making it particularly valuable for real-world pharmaceutical applications where new drugs continuously enter the market.

Abstract Large Language Models (LLMs) perform well in tabular prediction tasks with limited data, using their ability to understand instructions and learn from examples. However, their reliance on training data can perpetuate social biases, leading to unfair outcomes and disproportionately impacting underprivileged groups. Addressing these biases is critical as LLMs see wider adoption in tabular data tasks. Traditional bias mitigation strategies in machine learning, such as balancing datasets or applying fairness constraints, are less effective with LLMs. Our research explores whether bias in LLMs for tabular data classification can be mitigated. Through extensive experiments, we found that using LLMs in a zero-shot setting introduces bias, and in-context learning slightly reduces these disparities. Meanwhile, fine-tuning and retrieval augmented generation show limited effectiveness in bias mitigation. We introduced three instruction-based prompting strategies to enhance fairness: Fair Prompting , Generalised Prompting , and Descriptive Prompting . The results show that combining descriptive prompting with in-context learning, particularly the Equal Samples Across Demographics approach, consistently narrowed fairness gaps across demographic subgroups, and yielded accuracy gains ranging from 3.27% to 15.05% across multiple datasets, underscoring its potential as a promising strategy in the ongoing effort to mitigate bias in LLMs.

Large Language Models (LLMs) are increasingly deployed in web-based medical advice applications, offering scalable and accessible healthcare solutions. However, their outputs often reflect demographic biases, raising concerns about fairness and equity for vulnerable populations. In this work, we propose FairMed, a framework designed to mitigate biases in LLM-generated medical advice through fine-tuning and prompt engineering strategies. We evaluate FairMed using language-based and content-level metrics across demographic groups on publicly available (MedQA), synthetic (Synthea), and private (CBHS) datasets. Experimental results demonstrate consistent improvements over Llama3 - Med42, as well as over the zero-shot prompting baseline. For instance, in sentiment analysis for gender groups using MedQA, FairMed with Descriptive Prompting reduces the Statistical Parity Difference (SPD) from 0.0902 to 0.0658, improves the Disparate Impact Ratio from 1.1916 to 1.1566, and decreases the Kullback-Leibler Divergence from 0.0045 to 0.0024. Similarly, in directive language evaluation for gender groups using Synthea, SPD improves from 0.1056 to nearly zero, achieving near-perfect parity. On the CBHS dataset, FairMed with Descriptive Prompting increases Diagnostic Recommendation Divergence (DRD) for race groups from 0.9530 to 0.9848, indicating improved group-specific tailoring, while reducing the Action Disparity Index (ADI) from 0.0857 to 0.0469 and Referral Frequency Parity (RFP) from 0.0791 to 0.0511, reflecting enhanced fairness. These findings highlight FairMed's effectiveness in addressing demographic disparities and promoting equitable healthcare guidance through web technologies. This framework contributes to building trustworthy and inclusive systems for delivering medical advice by ensuring fairness in sensitive applications.

Link prediction is a crucial aspect of graph machine learning, with applications as diverse as disease prediction, social network recommendations, and drug discovery. It involves the prediction of potential new links between nodes within a network. Despite its importance, current models for link prediction exhibit notable limitations. Graph Convolutional Networks have shown high efficiency in link prediction across various datasets. However, they face significant challenges when applied to short-path networks and ego networks, resulting in poor performance. This issue represents a critical area of concern that our work seeks to address. This paper introduces the Node Centrality and Similarity Based Parameterised Model (NCSM), a novel method for link prediction tasks. NCSM uniquely integrates node centrality and similarity measures as edge features in a customised Graph Neural Network (GNN) layer, effectively leveraging the topological information of large networks. This model represents the first parameterised GNN-based link prediction model that considers topological information. The proposed model was evaluated on five benchmark graph datasets, each comprising thousands of nodes and edges. Experimental results highlight NCSM's superiority over existing state-of-the-art models like Graph Convolutional Networks and Variational Graph Autoencoder, as it outperforms them across various metrics and datasets. This exceptional performance can be attributed to NCSM's innovative integration of node centrality, similarity measures, and its efficient use of topological information.

Disease prediction has the potential to benefit stakeholders such as the government and health insurance companies. It can identify patients at risk of disease or health conditions. Clinicians can then take appropriate measures to avoid or minimize the risk and in turn, improve quality of care and avoid potential hospital admissions. Due to the recent advancement of tools and techniques for data analytics, disease risk prediction can leverage large amounts of semantic information, such as demographics, clinical diagnosis and measurements, health behaviours, laboratory results, prescriptions and care utilisation. In this regard, electronic health data can be a potential choice for developing disease prediction models. A significant number of such disease prediction models have been proposed in the literature over time utilizing large-scale electronic health databases, different methods, and healthcare variables. The goal of this comprehensive literature review was to discuss different risk prediction models that have been proposed based on electronic health data. Search terms were designed to find relevant research articles that utilized electronic health data to predict disease risks. Online scholarly databases were searched to retrieve results, which were then reviewed and compared in terms of the method used, disease type, and prediction accuracy. This paper provides a comprehensive review of the use of electronic health data for risk prediction models. A comparison of the results from different techniques for three frequently modelled diseases using electronic health data was also discussed in this study. In addition, the advantages and disadvantages of different risk prediction models, as well as their performance, were presented. Electronic health data have been widely used for disease prediction. A few modelling approaches show very high accuracy in predicting different diseases using such data. These modelling approaches have been used to inform the clinical decision process to achieve better outcomes.

In the modern financial market, market participants use big data analytics to gain valuable insight on historical market data for better decision making. Complying with the three vs (i.e., velocity, volume and variety) of big data, the financial market is considered as a complex system comprised of many interacting high-frequency traders those make decisions based on the relative strengths of these interactions. Researchers have put substantial scholarly input to deal with these anomalies. From the big data perspective, anomaly detection in financial data has widely been ignored despite many organisations store, process and disseminate financial market data for interested customers to assist them to make informed decision abd create competitive advantages. Considering the presence of anomalies in voluminous data from myriad data sources may generate catastrophic decision through misunderstandings of market behaviour. Therefore, in this study, we applied a standard set of anomaly detection techniques, used in big data based on nearest-neighbours, clustering and statistical approaches, to detect rare anomalies present within the historical daily trading information for five years (i.e., 2009--2013) for each stock listed on the Australian Security Exchange (ASX). We also measured the performance of these anomaly detection techniques using a number of metrics to highlight the best performing algorithm. The experimental results suggest that the LOF(Local Outlier Factor) and CMGOS(Clustering-based Multivariate Gaussian Outlier Score) are the best performing anomaly detection techniques.

Link prediction problem in network science has experienced extensive methodological improvements and simultaneously, spawned over numerous applications. In relation to evolutionary network analysis, different dynamic link prediction methods in network science not only support the prediction of future links but also assist in modelling network dynamics. The concept of constructing dynamic similarity metrics by considering the actor-level evolution of network structure and associated neighborhoods has been widely ignored for the purpose of dynamic link prediction. This study attempts to propose two dynamic similarity metrics for the purpose of dynamic link prediction in longitudinal networks through mining evolutionary information. These metrics consider the similarity between network structural and neighborhood changes over time incident to non-connected actor pairs. These metrics are then used as dynamic features in supervised link prediction model and performances are compared against two baseline static similarity metrics (i.e., AdamicAdar and Katz). Higher performance scores achieved by these features, examined in this study, exemplifies them as prospective candidates not only for dynamic link prediction task but also in understanding the growth pattern of dynamic networks.

The healthcare sector holds large amounts of semantically rich electronic data generated and used by different sections of the health care community. Data analytic techniques such as data mining and predictive modelling are being used to gain new insights into health care costs, performance and quality of care. In this context, social network analysis (SNA) has the unique ability to play a new role in exploring the context and situations that lead to efficient and effective healthcare. In this paper we describe a specific context of private healthcare in Australia and describe our SNA based approach (applied to health insurance claims) to understand the nature of collaboration among doctors treating hospital inpatients and explore the impact of collaboration on cost and quality of care. In particular, we use network analysis to (a) design collaboration models among surgeons, anaesthetists and assistants who work together while treating patients admitted for specific types of treatments (b) identify and extract specific types of network topologies that indicate the way doctors collaborate while treating patients and (c) analyse the impact of these topologies on cost and quality of care provided to those patients.

In this paper, we explore the effect of encircling behaviour on the topology of complex networks. We introduce the concept of topological encircling, which we define as an attacker making links to neighbours of a victim with the ultimate aim of undermining that victim. We introduce metrics to quantify topological encircling in complex networks, both at the network level and node pair (link) level. Using synthesized networks, we demonstrate that our measures are able to distinguish intentional topological encircling from preferential mixing. We discuss the potential utility of our measures and future research directions.

In this exploratory paper, we examine the evolution and engagement of an online community through a ten-year period. Data is collected from an online public discussion forum provided by a government-sponsored website specifically developed for community capacity building. We postulate that there are clear patterns of assortativity where similar actors engage in communication with each other over time. Results show that there is a clear pattern of networks losing their disassortative character in the early years followed by disassortative networks in the later years. The network-level results challenges government-level metrics of community-building success and suggests network analysis as an empirical avenue for understanding social processes involved in the very nature of community building.

In this paper, we introduce a measure to analyse the structural robustness of complex networks, which is specifically applicable in scenarios of targeted, sustained attacks. The measure is based on the changing size of the largest component as the network goes through disintegration. We argue that the measure can be used to quantify and compare the effectiveness of various attack strategies. Applying this measure, we confirm the result that scale-free networks are comparatively less vulnerable to random attacks and more vulnerable to targeted attacks. Then we analyse the robustness of a range of real world networks, and show that most real world networks are least robust to attacks based on betweenness of nodes. We also show that the robustness of some networks are more sensitive to the attack strategy compared to others, and given the disparity in the computational complexities of calculating various centrality measures, the robustness coefficient introduced can play a key role in choosing the attack and defence strategies for real world networks. While the measure is applicable to all types of complex networks, we clearly demonstrate its relevance to social network analysis.

Study of the dynamics of longitudinal networks has already attracted enormous research interest. Although dynamics of networks can be captured both at network-level and node / actor-level, the latter has gained less attention in current literature. By following a topological approach (i.e., static topology and dynamic topology) to analyze networks, this paper first proposes a research framework to capture actor-level dynamics for longitudinal networks. In static topology, Social Network Analysis (SNA) methods are applied on the aggregated network of entire data collection period. A smaller segment of network data that are accumulated in less time compared to the entire data collection period are used in dynamic typology for analysis purpose. This study further successfully compiles and applies this framework to the context of organizational crisis and project dynamics with the purpose to explore different level of actor-level dynamics at the different operational environment of these contexts over time. It is noticed that different level of actor-level dynamics are observed in the communication and collaboration network during the different facets of the organizations. In the context of organizational crisis, it is evident that during the 'crisis' period of operational running of organization, actors in the organizational email communication networks show higher level of actor-level dynamics compared to the 'normal' period. Less actor-level dynamics are observed during the 'final' phase of project life cycle, as found from the second context.

In this paper, we introduce a time-variant approach to degree centrality measure - time scale degree centrality (TSDC), which considers both presence and duration of links among actors within a network, whereas, the traditional degree centrality approach regards only the presence or absence of links. We illustrate the difference between traditional and time scale degree centrality measure by applying these two approaches to explore the impact of 'degree' attributes of doctor-patient network that evolves during patient hospitalization period on the hospital length of stay (LOS) both in macro- and micro-level. In macro-level, both the traditional and time-scale approaches to degree centrality can explain the relationship between the 'degree' attribute of doctor-patient network and LOS. However, at micro-level or small cluster level, TSDC provides better explanation while traditional degree centrality approach is impotent to explain the relationship between them.