21793 research outputs found
Sort by
Event graph optimization in RFI text documents using hierarchical reinforcement learning and human feedback
Get PDFRequests for Information (RFIs) are essential tools for facilitating communication among stakeholders in construction project management. They serve as formal inquiries to clarify ambiguities, resolve uncertainties, and enhance coordination between project teams, ultimately supporting effective project execution. RFIs play a critical role in maintaining workflow efficiency, reducing misinterpretations, and addressing unforeseen challenges that arise during the construction process. However, despite their importance, RFIs can sometimes highlight underlying systemic inefficiencies or anomalies that, if left unaddressed, contribute to cost overruns, schedule delays, and project quality issues.
Identifying the root causes of such anomalies within RFIs is crucial for mitigating risks and improving decision-making. Root Cause Analysis (RCA) is commonly employed to trace these issues back to their sources, allowing project managers and engineers to implement corrective measures. However, existing RCA approaches, many of which are based on event graphs, tend to rely on predefined rules or statistical correlations, which may not fully capture the dynamic and evolving nature of construction-related issues.
To address these challenges, this study proposes a novel approach that integrates human feedback-augmented reinforcement learning to enhance RCA for the event graphs of text-based RFIs. Our method leverages expert insights as a core component of the HRL loop in optimizing and improving the accuracy and reliability of causal and temporal graphs by effectively identifying and correcting errors within the graphs themselves. Importantly, our method is explicitly designed for structured RFI text data in which events have been manually pre-extracted as part of a preprocessing pipeline. Specifically, we employ hierarchical reinforcement learning (HRL) to systematically decompose the problem into multiple levels of decision-making, allowing for more structured learning and adaptation.
To validate our approach, we conduct experiments using the Causal-TimeBank dataset, a benchmark corpus annotated with explicit temporal and causal relationships. The reason for choosing this benchmark is largely due to the lack of publicly available RFIs with enriched text data and annotated causal/temporal events. Experimental results demonstrate that our method outperforms conventional RCA techniques by effectively identifying and correcting errors within causal and temporal graphs. The integration of human expertise ensures that the model remains adaptable to real-world complexities, enhancing its ability to capture nuanced relationships that might otherwise be overlooked by automated approaches. Ultimately, this work contributes to the advancement of intelligent RCA systems by combining human intuition with machine learning to create more robust, interpretable, and actionable root cause analyses in construction project management
Therapeutic Songwriting, Substance Use Disorder Recovery, and Trauma-Informed Care: A Philosophical Inquiry
Get PDFThis philosophical inquiry examined the use of therapeutic songwriting (TS) within a trauma-informed care (TIC) approach to support the processing of adverse childhood experiences within Substance Use Disorder (SUD) recovery. Drawing on the integral use of both outcome-oriented and experience-oriented songwriting methods (Bruscia, 2014) to support the processing of adverse childhood experiences (Kirkland, 2022), an argument was conceptualized and articulated by mapping specific songwriting elements (Baker, 2015) onto the specific principles of trauma-informed care. Relevant literature on SUD and SUD recovery, TIC, and therapeutic songwriting in SUD recovery was explicated to establish a foundation for the argument. Data analysis focused on specific songwriting elements and their potential to support key principles of TIC. Key findings demonstrated the potential for the utilization of a) the song components to support the TIC principles of safety and trustworthiness and transparency; b) the songwriting process to support the TIC principles of peer support, collaboration and mutuality, and cultural, historical and gender issues; c) the songwriting artifact to support the TIC principle of empowerment, voice and choice. Clinical, research, and educational implications are discussed along with limitations of the study
Senate Resolution on Open Science and Open Scholarship at Concordia University
Get PDFThis entry presents the Senate Resolution on Open Science and Open Scholarship, unanimously approved by Concordia University’s Senate on May 16, 2025. Building on Concordia’s longstanding leadership in open access, this resolution broadens the university’s commitment to a wide range of open practices—including open data, open code, open-source software, open educational resources, and citizen science.
Developed through an extensive, university-wide consultation process led by the Concordia Open Science Working Group, the resolution outlines a framework for advancing open practices across disciplines. It reflects three core aims: supporting implementation through training, infrastructure, and policy development; encouraging Concordians to engage with and adopt Open Science and Open Scholarship practices; and fostering recognition of these contributions within academic and research environments
An AI-Assisted Topic Model of the Media Literacy Research Literature
Get PDFMedia literacy, a vital field of research and educational practice, is attracting considerable scholarly attention, resulting in a burgeoning research literature. While numerous bibliometric studies have sought to capture the key features and themes of this body of literature, its rapid proliferation requires greater scalability and stronger capability to identify and characterize latent topics. In this study we address this gap by offering a computational bibliometric analysis of a corpus of 4,082 research documents on media literacy, spanning the period from 1985 to 2024. Through analysis of the documents’ metadata with natural language processing (NLP) using Latent Dirichlet Allocation (LDA) with Orange3, an open-access data mining software tool, we identify seven principal topics, each represented by a specific set of documents. The topics pertain to media publications and online content, critical thinking, youth behaviour, new media skills in education, news and misinformation, health (particularly among females), and communication strategies. We characterize these media literacy research topics with the assistance of a Large Language Model to generate a short synthetic description based on each topic’s top keywords. We complement our analysis with VOSviewer to produce co-citation maps of publication sources and authors to identify the disciplinary structure of the field, key ML authors, and their research contributions, which focus especially on media literacy education, digital media, behavioural issues, health impacts, and public perceptions
Response-Based Practice Informed Art Therapy for Non-Indigenous Practitioners Working with Indigenous Children Affected by Domestic Violence
Get PDFThe following research examines how non-Indigenous art therapy practitioners might ethically and effectively support Canadian Indigenous children who have experienced domestic violence
by integrating response-based practice into an art therapy intervention framework. While this study is grounded in the Canadian context, with appropriate cultural and contextual adaptation its findings may be relevant to art therapy practitioners in other settler-colonial nations working with Indigenous populations. Through a scoping review of relevant literature, this research identifies theoretical and practical alignments between response-based practice and various art therapy modalities, including client-centered, strengths-based, and trauma-informed approaches. Findings indicate that these approaches are compatible in their consideration of contextual factors, decentralization of pathology, promotion of client agency, and ethical mandate to pursue therapeutic work from an anti-oppressive stance, making their integration a promising direction for therapeutic work with this population. Tools like the Medicine Wheel of Resistance, and theoretical approaches such as the Expressive Therapies Continuum offer complementary methods for assessment and structuring therapeutic interventions, supporting both verbal and non-verbal forms of expression. However, the integration of these methods is underdeveloped in current literature and epistemological tensions present challenges. Additionally, the absence of direct consultation with Indigenous communities and children limits this research's applicability. Future studies should pursue collaborative, community-based research with Indigenous nations and communities to further develop and evaluate the effectiveness of integrative therapeutic models
Mapping the Internal Landscape: An Arts-Based Heuristic Self-Inquiry into Internal Family Systems-Guided Countertransference Response Art in Art Therapy Practice
Get PDFThis arts-based heuristic self-inquiry explores how Internal Family Systems (IFS) guided response art can support countertransference (CT) awareness and professional development in an art therapy intern. Drawing on the shared foundations of IFS and art therapy (including creativity, externalization, and a non-pathologizing stance) this study uses a parts-mapping exercise to guide post-session art making as a reflective tool. Over the course of clinical training, the researcher-participant created response artworks informed by the IFS framework, followed by journal reflections. The data was analyzed through Moustakas’s six phases of heuristic inquiry, with particular focus on the illumination and explication of internal processes. Findings suggest that IFS-guided response art offers a valuable method for deepening self-awareness, enhancing CT management, and fostering professional growth in novice art therapists. This study contributes to the emerging literature on integrative models in art therapy and highlights the importance of creative methods in early therapist development
Online Steiner Cover Problems in Hypergraphs
Get PDFThe online Steiner cover problem in hypergraphs (\treePN) is a generalization of the online Steiner tree problem in graphs.
An edge-weighted hypergraph is given offline and a set of terminal vertices is requested sequentially online.
Upon receiving each request an algorithm for the problem must buy some edges which connect to the previous solution.
The solution after satisfying the request is then the union over all edges bought up to that point, .
The goal is to minimize the total cost of the solution to connect the requests, i.e., for let , then we want to minimize .
The generalized \treePN (\forestPN) is a generalization of both the \treePN and the Steiner forest problem in graphs.
Again, we are given an edge-weighted hypergraph offline, but instead of a set of terminals as the online portion of the input we are given a set of terminal pairs .
Upon receiving the request pair an algorithm for this second problem must buy some set of edges which connect the terminals from .
We define the instantaneous solution after connecting request as before, so and the final solution is again denoted for a request sequence of size .
The worst-case performance of an online algorithm is measured by the competitive ratio, which is the ratio between the cost of a solution obtained by the online algorithm to that of an optimal offline solution.
Besides some simpler preliminary results, we obtain a lower bound on the competitive ratio for \treePN (which also applies to \forestPN) of , where is the rank of the hypergraph, and a matching upper bound for the simple algorithm.
For \forestPN we show that the simple algorithm is -competitive and provide another algorithm, called , which achieves a competitive ratio of
A Design and Implementation of Learned Index for Processing Multi-Dimensional Queries Over Relational Data
No full textWe study the performance evaluation and analysis of Flood, a learned index designed to process multi-dimensional queries over relational data. Unlike traditional indexing methods such as KD-Trees and R-Trees, which rely on static partitioning strategies, Flood leverages machine learning techniques to dynamically adapt its grid structure and data layout based on data distribution and query workloads. We identify and evaluate the core components of the Flood framework—including grid-based partitioning, learned layout optimization, and refinement steps—and assess the impact of each component on overall system performance. We compare Flood’s efficiency and scalability against baselines such as KD-Trees, Z-order indexing, and brute-force scan across multiple datasets and query workloads. Our experiments reveal that scan emerges as the dominant bottleneck, with layout tuning and component configuration introducing significant overhead. To gain a deeper understanding of Flood and explore opportunities for improvement, we developed a modular prototype implementation from scratch. This modular design enabled a systematic, in-depth performance study by isolating components and allowing for alternative configurations. We also refined the cost model calibration and proposed a new optimization strategy for guiding layout selection. Our work contributes to more effective configuration and tuning of learned indexes by offering insights into the trade-offs, limitations, and opportunities of using learned indexes as an alternative or complementary solution for supporting multi-dimensional queries in relational database systems
Concilier les fonctions minorisante et hédoniste des culturèmes irlandais dans la traduction française du roman An Irish Country Girl de Patrick Taylor au moyen du postulat traductif
No full textLe présent mémoire commente la mise en application du postulat traductif visant à produire un texte littéraire agréable à lire (fonction hédoniste) tout en préservant les marqueurs de la culture irlandaise (fonction minorisante). Concept fondé sur l’approche fonctionnaliste, le postulat traductif propose de dégager une hypothèse de traduction à partir des fonctions du texte à traduire et de la décliner en priorités afin d’orienter la démarche traductionnelle. Forte de cette approche, la présente traduction commentée du roman An Irish Country Girl de l’auteur irlando-canadien Patrick Taylor concilie la fonction minorisante et la fonction hédoniste en mettant à profit une troisième fonction, qui a une finalité informative. Grâce à celle-ci, il devient possible d’appliquer différentes stratégies visant à mettre en évidence les particularités linguistiques et culturelles de la minorité irlandaise sans compromettre le plaisir de la lecture au moyen de procédés péritextuel et intratextuels
Three Essays in Mental Health Economics: Education and Relationship and Career Stability
No full textThis dissertation investigates the long-term economic consequences of mental health, tracing its role across three interrelated domains: educational attainment, relationship stability, and career stability. Using panel data from the National Longitudinal Study of Adolescent to Adult Health (Add Health), the dissertation applies structural modeling techniques to capture the multidimensional nature of mental health and its evolving impact across the life course.
The first chapter addresses a gap in the economics of education literature by examining how adolescent mental health influences high school completion and college enrollment. This study uses a two-step framework combining exploratory factor analysis with Generalized Structural Equation Modeling (GSEM) and logistic regression. Latent variables for mental and physical health are constructed using indicators such as ADHD, anxiety, depression, substance use, subjective health, and BMI. Results show that mental health significantly, though modestly, affects educational outcomes, with a stronger effect on college entry than on high school completion. A shift from poor to excellent mental health is associated with a 6.9 percentage point increase in college enrollment and a 4.2-point rise in high school completion.
The second chapter extends the analysis to intimate partnerships. Using a Cross-Lagged Panel Model (CLPM), the chapter demonstrates a significant bidirectional relationship between mental health and relationship stability. Individuals with better mental health experience more stable relationships, which in turn improve mental well-being. Moving from poor to excellent mental health predicts a 6.7-point rise in relationship stability, while stable relationships enhance mental health by 5.4 points.
The third chapter explores how mental health affects job stability using a mediated Structural Equation Model. Latent constructs for job stability, mental and physical health, estimated in an earlier stage, are used to examine both direct and indirect effects through mediating pathways such as prior relationship stability and physical health. A full-range improvement in mental health increases predicted job stability by 20.3 percentage points