30472 research outputs found
Sort by
Trust models go to the web: learning how to trust strangers
Get PDFWe study emerging traits of interpersonal and social trust in online social networks of needs (OSNNs), where trust interactions start online and evolve into in-person meetings.
We present a lightweight web scraping solution to harness data from online social networks; thanks to it we were able to monitor a nation-wide portal for childcare and see the evolution of online reviews from both families and carers.
We analysed the data by first considering topological information to test centrality metrics as proxies for trustworthiness.
Next, we focused on features/profile analysis and tested the Castelfranchi–Falcone trust model from psychology (CF-T), fitting it to online reviews of childcare services.
Even though such reviews are relatively scarce and seemingly skewed, we feature-engineered the CF-T model to predict the evolution of reviews, treated as proxies for trust.
By aggregating CF-T scores at the regional level, we discovered a strong correlation with per capita GDP, which suggests that high levels of trust in social networks of needs reflect social capital
Interventions to reduce empathy-based stress and enhance compassionate care in mental health wards: a systematic review
Get PDFBackground: Mental health wards are an important healthcare context with the potential to positively impact patient trajectories. Compassionate care in these wards is important, and can be impacted by staff levels of empathy-based stress (compassion fatigue, burnout and secondary trauma). It is important to consider the evidence-base for mental health ward interventions to improve compassionate care for patients and to reduce empathy-based stress for staff.
Methods: A systematic review was conducted of robust evaluations of mental health ward interventions designed to improve compassionate care and/or reduce staff empathy-based stress, with the aim of synthesising interventional evidence on these interventions’ effectiveness, implementation and acceptability. Programme theory papers, outcome evaluations (RCTs and non-RCTs), economic evaluations and process evaluations were included. A meta-integration of intervention content, effectiveness and influence of contextual factors on implementation and acceptability was performed.
Results: 18 eligible study reports of 11 interventions were identified. Interventions were multi-level, and aimed to increase staff resources rather than decrease staff demands. Staff training interventions were most evaluated, with mixed evidence for effectiveness. Other approaches included changes to ward approach, environment, use of participatory action research methods and peer-review programmes. There was no clear evidence for a particular intervention type. Two interventions showed evidence of iatrogenic harm. Equity harms and economic effects were not well-evaluated. Mechanisms of change were under-theorised and lacked clear logic models. Patient and public involvement was sporadic.
Conclusions: Current interventions are being offered without a clear evidence-base or guiding model, and risk harming staff. Multi-level interventions using clearer logic models which tackle both job demands and resources are recommended. A model of implementation factors which may help interventions to succeed is proposed. More high-quality controlled intervention studies, considering contextual and process factors, and incorporating co-production, are needed, especially given the risk of iatrogenic harm
Near-factorizations of dihedral groups
Get PDFWe investigate near-factorizations of nonabelian groups, concentrating on dihedral groups. We show that some known constructions of near-factorizations in dihedral groups yield equivalent near-factorizations. In fact, there are very few known examples of nonequivalent near-factorizations in dihedral or other nonabelian groups; we provide some new examples with the aid of the computer. We also analyse a construction for near-factorizations in dihedral groups from near-factorizations in cyclic groups, due to Pêcher, and we investigate when nonequivalent near-factorizations can be obtained by this method
Closed form representations for the compactly supported radial basis functions of Buhmann, Wendland and Wu
No full textThe original compactly supported radial basis functions of Wendland and Wu have a polynomial form and are constructed using a two-step dimension walk strategy. Focussing on the Wendland functions, Schaback proposed a one-step dimension walk which is shown to recover the original Wendland functions at every second step but also introduces new examples, the so-called missing Wendland functions at the intermediate steps. In a recent paper by Huang et at., the analogue of Schaback's work is presented for the Wu functions and so deliver the so-called missing Wu functions. The original and missing Wendland functions belong to a much wider class proposed by Buhmann. The classical Buhmann functions, which are related to thin-plate spline radial basis functions, also belong to this much wider class. The theme uniting the classical Buhmann functions and the missing Wendland/Wu functions is that they are
non-polynomial and closed form expressions are not known for all of them. In this paper we revisit these functions and show how closed form representations can be given using direct techniques. The results for the classical Buhmann and Wu functions are new and the resulting expressions for the missing Wendland functions improve on those given iby the first author and so their implementation should be more straightforward
IV. Arrested feedback loops: student experience and teachers’ evaluation
No full textStudent-centred approaches to teaching and learning are premised on a feedback loop moving between learning objectives, student experience, and student performance. The latter is relied upon in the evaluation of the achievement of the learning objectives through concrete measurable outcomes. Each step in this loop is critical for ensuring that pedagogic practice successfully aligns learning objectives with the needs and abilities of a specific body of students, but what happens when the flow of this loop is interrupted? This paper considers a moment in an undergraduate classroom where the flow of feedback loop comes to a halt. It asks how we can make sense of discordance between students’ reported experiences and teachers’ evaluations. In so doing, we explore the nuanced relationship between experience and evaluation not as nodes in a seamless loop, but as sites where the relationship amongst the different stakeholders (students-teachers-institutional evaluative criteria) is renegotiated and potentially reconfigured
A multi-label visualisation approach for malware behaviour analysis
Get PDFModern malware evolves continuously, posing persistent challenges to cybersecurity. Conventional classification approaches typically group malware by its primary objective, emphasising dominant behaviours while overlooking the complex and overlapping strategies common in real-world attacks. Here we present DECODE (DEep Classification Of Dynamic Exploits), a proportional multilabel, context-aware framework that combines object detection, explainable artificial intelligence (XAI), and agent-based large language models (LLMs) to deliver interpretable and comprehensive malware analysis. DECODE introduces the first object detection dataset specifically for malware classification, generated through an automated annotation pipeline that removes the need for manual labelling and remains effective even for visually indistinguishable malware features. To improve
attribution reliability, we extend Gradient-weighted Class Activation Mapping (Grad-CAM) with a Bayesian formulation, enabling uncertainty-aware visualisation of discriminative regions linked
to multiple categories. The regions identified through object detection are subsequently mapped to their corresponding API call sequences and interpreted via a multi-agent reasoning module,
which incorporates critique-and-verification loops to reduce hallucinations and bias. Experimental evaluation shows multi-label and binary classification accuracies of 0.8513 and 0.9380, respectively, outperforming conventional deep learning baselines. By combining visual localisation, proportional multi-label scoring, and human-readable behavioural narratives, DECODE enables malware to be classified not only by intended impact but also by fine-grained structural and behavioural traits, offering a richer understanding of complex threats
Recovery and rehabilitation in mental health: historical perspectives
Get PDFRecovery and rehabilitation are highly charged terms in contemporary mental health, with their meanings and implications contested by professionals and survivors alike. A loose ‘recovery movement’ with radical reformist aims, which emerged across Britain and the United States in the later decades of the 20th century, has, some argue, been co-opted by ‘neoliberal’ political and clinical interests. Most of these narratives begin with the emergence of the recovery movement as a result of service-user/peer activism in the 1970s: few consider the longer history of ideas and practices of recovery and rehabilitation. In turn, recovery and rehabilitation have been strangely marginal in the works of historians of psychiatry, madness, and mental health. This article traces some of the key literatures and concepts in relation to recovery and rehabilitation, and introduces this special issue of History of the Human Sciences, with contributions on Britain and the United States from the late 19th century to the turn of the 21st century
On What Mutters: The Unnamable Subject of Practical Reason - Finnis..Raz...Beckett..Cha
Get PDFWhy read Beckett’s The Unnamable and Cha’s Dictée alongside the philosophy of practical reason? Such a reading does not seek to reject practical reason, or to suggest that it can be replaced by thinking that draws on literary texts. The point is to make practical reason strange through its encounter with what mutters or murmurs, resists a name and, at the same time, obligates speech. If we address this peculiar concern, then we might be able to engage with the troubling sense of non being that haunts the subject of practical reason; and this theme will be pursued through texts by John Finnis and Joseph Raz. Cha and Beckett compel us to think about the constitution of the subject as a being capable of bearing its own non being. In so doing, we encounter a radical indeterminacy that is inseparable from the productive work of the unconscious; its metaphorization of the inherent flaws in being. This ‘is’ what is unnamable. In terms borrowed from Abraham and Torok we can see this figuring of a constitutive flaw as an endless, ongoing project of introjection. To make use of a psychoanalytic term like introjection is not to suggest that psychoanalysis offers a symbolic key that will unlock the secrets of the texts with which we will be dealing. Rather, we are in pursuit of terms that will allow us to relate themes together in a productive way. To put this in Beckettian terms, we will be addressing an endless – and endlessly interrupted- desire to start again. To read again. Read differentl
Data sources:
Access Repository Dashboard
Do you manage Birkbeck Institutional Research Online? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!