TARA

Click the image to view all open access publications by TCD authors or view by academic unit by clicking the link on the left menu bar.

TCD research that has been funded by the EC is available here and in the EC's open access portal at http://www.openaire.eu.

Trinity Irish Art Research Centre Image Collections.

The complete archive of the Journal of the Statistical & Social Inquiry Society of Ireland is available on Open Access via TARA. Click the image to access.

Céad Míle Fáilte...

TARA is an open access repository, which means that the full text of the work deposited here is freely accessible to the world via the web. Access is restricted only in unavoidable instances, for example where publisher copyright restrictions prevail. However over 90% of scholarly publishers worldwide now allow some version of the documents they publish to be made available in a repository such as this.

Recent Submissions

FAIR Marine Data Workflows for Policy: Unifying Seabed Integrity and Connectivity in Irish SACs via EDITO and DestinE
(2026-03-14) Crowley, Quentin
Making marine geospatial data Findable, Accessible, Interoperable, and Reusable (FAIR) remains challenging for researchers and policy implementors, particularly in integrating geological and biological datasets for Special Areas of Conservation (SACs) management. This contribution shares experiences developing domain-specific FAIR workflows for west coast Ireland SACs (Porcupine Seabight, Belgica Mound, Inisheer Island), harmonizing INFOMAR multibeam data, EMODnet Geology, OBIS biodiversity, and Copernicus currents via the European Digital Twin Ocean (EDITO) and Destination Earth (DestinE) platforms (and others). Seabed integrity metrics (e.g., Bedrock Suitability Index information) and substrate maps (85% accuracy, Random Forest classification) will be processed on available platforms, e.g., EDITO and DestinE HPC, post-QC for best possible and valid geometries and INSPIRE compliance. Biodiversity connectivity matrices (previous published work and code from the coastalNet R package will be cited and explored), pairwise probabilities e.g., 0.35 Belgica-to-Porcupine) overlay oceanographic simulations (e.g., ESRI EMUs), deposited as interoperable WMS layers on Figshare DOIs with plain-language metadata and APIs. Specific challenges include integrating "dark" datasets and bridging technical-policy gaps; solutions involved AI-driven summarization, automated versioning, and user-centric pilots (e.g., co-design workshops, tracking download rates, policy citations). Additional challenges include alignment with MSFD thresholds (>25% degraded seabeds) and OSPAR goals fostered adoption, with sensitivity analyses (low BSI reduces connectivity 20-40%) potentially useful for informing trawling vignettes and conservation and restoration efforts (reefs on BSI>0.7). This approach respects ocean science needs while promoting cross-disciplinary understanding and reuse (e.g., hydrology via sediment mobility), demonstrating cultural shifts through stakeholder panels and GDPR-compliant training toolkits. Outcomes advance RDA ESES goals by scaling FAIR practices for real-time AI dashboards, inviting dialogue on community-driven refinement.
Understanding the Nature of Engagement with Executive Managers to Create and Sustain Strategic Alignment: The Case of Network Utility Organisations in Ireland
(Trinity College Dublin. School of Business. Discipline of Business & Administrative Studies, 2026) Kuzucu, Cenk; McDonagh, Joseph
Strategic alignment between business and information systems (IS) strategies remains a persistent challenge for organisations, particularly in complex, regulated, and resource-constrained settings of Network Utility Organisations in Ireland. This thesis explains how alignment is created and sustained over time by developing an empirically grounded account that integrates both process-oriented and practice-oriented perspectives. Adopting a relativist ontology and interpretivist/constructivist epistemology, the study employs a qualitative, longitudinal multi-case design, combining narrative strategy, processual inquiry, thematic analysis and grounded theory strategy. The thesis offers several key contributions. First, it operationalises the "how" of alignment through the Strategic Alignment Process Framework, mapping the phases, stages and IS artefacts that structure alignment work. Second, it reconceptualises strategic alignment as a dynamic, cyclical construct with the Strategic Alignment Lifecycle Loop, formalising feedbacks between strategy formulation, delivery execution and operations, and ongoing adaptation. Third, the research develops a mechanism-level account of social alignment via the Executive Manager Engagement Cycle, detailing the routines and set of novel attributes that make executive engagement purposeful and effective. Fourth, the study proposes Financial Alignment as a distinct alignment type, demonstrating that financial scrutiny and value justification are central, recurring mechanisms in alignment practice. Fifth, this study introduces a novel theoretical framework that synthesises the co-evolutionary perspective with the enterprise architecture view. The thesis seeks to demystify the strategic alignment by situating Strategic Alignment Process Framework at the epicentre of Strategic Alignment Lifecycle Loop and Executive Manager Engagement Cycle. Within this construct, alignment types act as intervening mechanisms that initiate, shape, and ensure the sustainment of strategic coherence across successive cycles, offering a deeper understanding of how alignment is created and sustained over time. The findings offer theoretical, empirical and practical value, providing actionable frameworks, engagement toolkits, and artefact templates that organisations can adapt to structure, diagnose, and improve alignment. The research lays a foundation for future studies to test and refine these models in other contexts and contributes to bridging the gap between theory and practice in strategic alignment.
Short-Horizon Data-Driven Joint Forecasting of Wind Speed and Direction for Future Aware Wind Farm Control
(2026) Alazhare, Abdulbaset; Howland, Michael; Fitzgerald, Breiffni
Preview-based wind-farm control depends on accurate short-horizon estimates of wind speed and direction. Yet at five-minute resolution, practical models must be low-latency, and must outperform persistence modelling. This paper presents a two-step, control-oriented data-driven forecasting model for one-step-ahead prediction of the wind vector components. In step 1, the study compares a Long short-term memory network (LSTM), a 1-D convolutional neural network (CNN), and a causal temporal convolutional network (TCN) against a one-step persistence baseline, under a unified training and evaluation protocol. In step 2, the study retains the best-performing architecture and enhances the input representation using variational mode decomposition (VMD) and mutual-information (MI) lag analysis, yielding a VMD–MI–TCN model. The study uses three consecutive years of five-minute wind speed and direction from NREL’s Wind Resource. Forecasts are evaluated on a held-out test set using wind-speed errors, circular wind-direction errors, and error-based skill scores relative to persistence. Overall, the proposed workflow provides a basis for constructing and benchmarking lightweight wind speed and wind direction forecasting models suitable for real-time control integration.
Robust Multimodal Turn-Taking for Human-Machine Interaction
(Trinity College Dublin. School of Engineering. Discipline of Electronic & Elect. Engineering, 2026) O'Connor Russell, Samuel Arthur; Harte, Naomi; Research Ireland; ADAPT Centre
During conversation, humans rapidly switch from speaking to listening and vice versa. This process is called turn-taking. Whilst second nature for humans, turn-taking presents a major challenge for emerging technologies such as voice assistants. At present, the majority of systems rely on detecting the silence after a speaking turn. This fails to replicate the fast-paced nature of human-human interaction. In recent years, predictive turn-taking models (PTTMs) have been proposed. Inspired by human turn-taking, PTTMs are neural networks which continuously predict whether or not a speaker change will occur in the near future. These models form an active area of research, and they have grown in sophistication and performance. This thesis addresses a number of under-explored areas in PTTM development. Currently, a major obstacle is PTTMs rely upon manual transcription of conversations; a timely and costly process. The thesis therefore offers a comprehensive overview of automatic speech recognition (ASR) for automated conversational speech transcription. Insights include a gap in the performance of commercial and open-source ASR on the transcription of non-lexical aspects of speech. We then explore PTTM training with ASR transcription, finding no difference in the performance of PTTMs trained using word alignments from ASR and manually transcribed interactions. This enables PTTM training on larger, more diverse multimodal datasets for which manual transcription is typically unavailable. PTTMs rely on speech alone to make their predictions, but turn human turn-taking is richly multimodal. The inclusion of visual cues such as gaze in PTTM has only been explored in a single, limited study. This major gap in the literature is addressed in this thesis by introducing the multimodal voice activity projection (MM-VAP) model, one of the first PTTMs to consider the inclusion of visual cues such as gaze, head pose and facial expression. We find that MM-VAP significantly outperforms the state-of-the-art audio-only approach: MM-VAP achieves an 83% balanced accuracy on the hold/shift prediction task, compared with 79% balanced accuracy in the audio-only state-of-the-art at the same noise level. Our evaluation is conducted across a comprehensive range of events, such as detecting the next speaker during silence and before overlapping speech. A further gap in the literature is that the performance of PTTMs in noise has not been considered, yet models will inevitably encounter noise once deployed. We address this in the thesis by adding environmental noises such as babble and music noise to the Candor videoconferencing corpus. We find that our MM-VAP model is not inherently robust to noise, with performance falling from 83% to 54% balanced accuracy in +10 dB music noise. However, we find that when we adapt the training of the model to include examples of noise, MM-VAP outperforms the audio-only state-of-the-art audio-only model by a wide margin, at 75% balanced accuracy in +10 dB speech noise compared with 64% balanced accuracy audio-only state-of-the-art at the same noise level. We demonstrate that the performance increase arises from the exploitation of visual cues in the MM-VAP model. However, we find that the performance increase does not always generalise to new sources of noise, highlighting the importance of the training process in PTTM development. The thesis also explores the role of multimodal cues within the speech signal. Recent PTTMs leverage self-supervised speech representations (S3Rs). These are learned representations as they capture prosodic, lexical and semantic aspects of the speech signal. However, it is difficult to establish which aspects of speech are exploited for prediction. We propose a vocoder-based methodology to selectively control the amount of prosodic and lexical information in speech. We find that the voice activity projection (VAP) model, an S3R-based turn-taking model from the literature, utilises both prosodic and lexical cues for prediction. However, when one feature is corrupted, the model can flexibly utilise the other without further training. A notable finding is that when speech is replaced with unintelligible noise following the prosodic contour of speech, performance remains above chance, at 69% balanced accuracy. This demonstrates that the prosodic contour alone is a powerful predictor of turn-taking patterns. We expand on this analysis by comparing the audio VAP model with large language model (LLM) based turn-taking models, which rely exclusively on text. We find that the LLM based models have a high number of false positive end-of-turn predictions. The audio-only turn-taking model does not suffer from the same errors, suggesting prosody plays a key role in turn-taking prediction. We therefore propose the Audio TurnGPT model, an LLM based turn-taking model which incorporates acoustic cues. We find that the model significantly improves prediction performance (90% vs. 85% balanced accuracy). The thesis therefore demonstrates the complementary, additive role of multimodal cues in turn-taking prediction. The best-performing models exploit all cues, but there is a degree of redundancy, as individual cues can perform as turn-taking predictors. Multimodality not only improves performance but also increases robustness to noise, and means models retain performance even when cues are removed. The analyses, methodology, and models introduced in this thesis represent significant advances in the predictive turn-taking literature. It is hoped that this thesis leads to an increased awareness of multimodality when working human interaction data. All modalities that are available to interlocutors in a dataset should be considered, as should their interactions with another and their influence on unfolding interactions. We also offer a discussion on future work and the implications of our findings.
Oral Health and Function in Older Adults with Intellectual Disabilities: Measurement Innovation for Epidemiological and Population Health Research
(Trinity College Dublin. School of Dental Sciences. Discipline of Dental Science, 2026) Byrne, Katrina; MacGiolla Phadraig, Caoimhin; Dublin Dental University Hospital; Trinity College Dublin
Summary Background Oral health is increasingly recognised as integral to systemic wellbeing, functional capacity and survival in ageing populations. However, adults with intellectual disabilities remain largely excluded from oral health surveillance and longitudinal epidemiological research. As a result, their oral health status and its interaction with multimorbidity, frailty and premature mortality remain insufficiently understood. This exclusion reflects both methodological barriers and structural inequities in health research and policy. Aim The overarching aim of this thesis was to generate disability-inclusive oral health evidence by developing and validating a structured clinical oral health assessment tool and embedding it within a nationally representative longitudinal ageing cohort to examine oral health status, oral�systemic associations and mortality risk among older adults with intellectual disabilities in Ireland. Methods The thesis adopted a post-positivist epidemiological framework informed by critical realism and an equity-oriented lens. It comprised three interconnected components. First, scoping review synthesised international evidence on oral health and pneumonia in adults with intellectual and developmental disabilities. Second, the Modified Oral Status Survey Tool (MOSST) was developed and psychometrically evaluated to enable structured clinical oral health assessment by trained non-dental clinicians. Content validity was established through international expert review and public and patient involvement. Inter- and intra-rater reliability exceeded recommended epidemiological thresholds (� > 0.85), and feasibility testing demonstrated high acceptability within field conditions. Third, MOSST was embedded within Wave 5 of the Intellectual Disability Supplement to The Irish Longitudinal Study on Ageing (IDS-TILDA), enabling cross-sectional and longitudinal analyses of oral health, systemic conditions and mortality. Multivariable regression models examined associations between oral indicators and systemic outcomes. Cox proportional hazards models assessed associations between tooth loss and all-cause and respiratory-related mortality. Results The scoping review identified biological plausibility for oro-respiratory pathways but highlighted a fragmented and methodologically limited evidence base. MOSST demonstrated strong content validity, high reliability and feasibility within a national ageing cohort. Population-level analyses revealed a substantial burden of untreated caries, compromised gum condition, reduced functional dentition and incomplete prosthetic rehabilitation among older adults with intellectual disabilities. Clear demographic and contextual patterning were observed by age, level of intellectual disability and residence type. In adjusted multivariable models, absence of functional dentition was significantly associated with constipation and osteoporosis, while indicators of poor oral cleanliness and disease burden were associated with respiratory illness. Longitudinal survival analyses demonstrated that tooth loss and absence of functional dentition were independently associated with increased risk of both all-cause and respiratory-related mortality, even after adjustment for age, gender, level of intellectual disability and multimorbidity. Discussion These findings demonstrate that oral health among older adults with intellectual disabilities is neither inevitable nor clinically peripheral but structurally patterned and epidemiologically consequential. By embedding a validated disability-adapted oral health assessment within a longitudinal cohort, this thesis advances inclusive epidemiological infrastructure and enables examination of oral-systemic and oral-mortality pathways previously absent from disability research. The results support conceptualisation of oral health as a modifiable determinant within multimorbidity and frailty trajectories and highlight the role of structural inequities in shaping oral health outcomes. Conclusions This thesis makes three principal contributions: (1) development and validation of a disability-inclusive oral health assessment tool suitable for large-scale surveillance; (2) generation of the first nationally representative, clinically assessed oral health dataset for older adults with intellectual disabilities in Ireland; and (3) empirical demonstration of oral�systemic and oral�mortality associations within a longitudinal disability cohort. The findings support integration of oral health into ageing research, public health monitoring and rights-based disability policy, reframing oral health inequity as a structural issue rather than an inherent feature of disability.