Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which (...) they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy. (shrink)

Machine behavior that is based on learning algorithms can be significantly influenced by the exposure to data of different qualities. Up to now, those qualities are solely measured in technical terms, but not in ethical ones, despite the significant role of training and annotation data in supervised machine learning. This is the first study to fill this gap by describing new dimensions of data quality for supervised machine learning applications. Based on the rationale that different social (...) and psychological backgrounds of individuals correlate in practice with different modes of human–computer-interaction, the paper describes from an ethical perspective how varying qualities of behavioral data that individuals leave behind while using digital technologies have socially relevant ramification for the development of machine learning applications. The specific objective of this study is to describe how training data can be selected according to ethical assessments of the behavior it originates from, establishing an innovative filter regime to transition from the big data rationale n = all to a more selective way of processing data for training sets in machine learning. The overarching aim of this research is to promote methods for achieving beneficial machine learning applications that could be widely useful for industry as well as academia. (shrink)

The annotated dataset is an essential requirement to develop an artificial intelligence system effectively and expect the generalization of the predictive models and to avoid overfitting. Lack of the training data is a big barrier so that AI systems can broaden in several domains which have no or missing training data. Building these datasets is a tedious and expensive task and depends on the domains and languages. This is especially a big challenge for low-resource languages. In (...) this paper, we experiment and evaluate many various approaches on sentiment analysis problems so that they can still obtain high performances under limited training data. This paper uses the preprocessing techniques to clean and normalize the data and generate the new samples from the limited training dataset based on many text augmentation techniques such as lexicon substitution, sentence shuffling, back translation, syntax-tree transformation, and embedding mixup. Several experiments have been performed for both well-known machine learning-based classifiers and deep learning models. We compare, analyze, and evaluate the results to indicate the advantage and disadvantage points of the techniques for each approach. The experimental results show that the data augmentation techniques enhance the accuracy of the predictive models; this promises that smart systems can be applied widely in several domains under limited training data. (shrink)

Bowers et al. argue that deep neural networks (DNNs) are poor models of biological vision because they often learn to rival human accuracy by relying on strategies that differ markedly from those of humans. We show that this problem is worsening as DNNs are becoming larger-scale and increasingly more accurate, and prescribe methods for building DNNs that can reliably model biological vision.

Metacognitive training (MCT) was developed to promote awareness of reasoning biases among patients with schizophrenia. While MCT has been translated into 31 languages, most MCT studies were conducted in Europe, including newer evidence recommending an individualized approach of delivery. As reasoning biases covered in MCT are separable processes and are associated with different symptoms, testing the effect of selected MCT modules would help to develop a targeted and cost-effective intervention for specific symptoms and associated mechanisms. This study tested the (...) efficacy of a 4-session metacognitive training for delusions, MCTd (in Traditional Chinese with cultural adaptations, provided individually), as an adjunct to antipsychotics in reducing severity and conviction of delusions, jumping to conclusions bias and belief inflexibility. Forty-four patients with delusions were randomized into the MCTd or the wait-list control condition. Patients on wait-list received the same MCTd after four weeks of treatment as usual. Assessment interviews took place before and after the treatment, and at 4-week follow-up. There was an additional baseline assessment for the controls. Jumping to conclusions and belief flexibility were measured by the beads tasks and the Maudsley Assessment of Delusions Scale. Attendance rate of the MCTd was satisfactory (84.5%). Compared to treatment as usual, there was a greater reduction in psychotic symptoms, delusional severity and conviction following MCTd. There was a large treatment effect size in improvement in belief flexibility. Improvement in reaction to hypothetical contradiction predicted treatment effect in positive symptoms and delusions. Jumping to conclusions bias was reduced following MCTd, although the treatment effect was not significantly larger than TAU. Our results support the use of process-based interventions that target psychological mechanisms underlying specific psychotic symptoms as adjuncts to more conventional approach. (shrink)

Cognitive control processes, such as updating task-relevant information while switching between multiple tasks, are substantially impaired in older adults. However, it has also been shown that these cognitive control processes can be improved by training interventions, e.g., by training in task switching. Here, we applied an event-related potential approach to identify whether a cognitive training improves task-preparatory processes such as updating of relevant task goals. To do so, we applied a pretest-training-posttest design with eight training (...) sessions. Two groups of older adults were either trained in task switching or in performing single tasks and we compared their performance to a group of untrained younger adults. To foster cue updating in the treatment group, we applied a cue-based switching task in which the two task cues were randomly selected prior to target presentation so that participants had time to prepare for the upcoming task. In contrast, the control group also received task cues but those were redundant as only one task had to be performed. We also examined whether training in cue updating during task switching can be transferred to a similar cognitive control task measuring updating of context information, namely a modified version of the AX-Continuous Performance Task. The results revealed training-specific improvements in task switching, that is, a larger improvement in blocks requiring switching in comparison to single tasks at the behavioral level. In addition, training specific-effects were also found at the neuronal level. Older adults trained in cue updating while switching showed a reduction in mixing costs in the cue-related P3, indicating an improvement in preparatory updating processes. Additionally, P3 topography changed with training from a very broad to a parietally focused scalp distribution similar to the one found in younger adults. However, we did not obtain training-specific improvements in context updating in the AX-CPT neither at the behavioral level nor at the neuronal level. Results are discussed in the context of the ongoing debate on whether transfer of cognitive training improvements is possible. (shrink)

In Australia research projects proposing the use of linked data require approval by a Human Research Ethics Committee . A sound evaluation of the ethical issues involved requires understanding of the basic mechanics of data linkage, the associated benefits and risks, and the legal context in which it occurs. The rapidly increasing number of research projects utilising linked data in Australia has led to an urgent need for enhanced capacity of HRECs to review research applications involving this (...) emerging research methodology. The training described in this article was designed to respond to an identified need among the data linkage units in the Australian Population Health Research Network and HREC members in Australia. (shrink)

This paper presents social data and knowledge collective intelligence platform for TRaining Ethical AI Models (STREAM) to address the challenge of aligning AI models with human moral values, and to provide ethics datasets and knowledge bases to help promote AI models “follow good advice as naturally as a stream follows its course”. By creating a comprehensive and representative platform that accurately mirrors the moral judgments of diverse groups including humans and AIs, we hope to effectively portray cultural and (...) group variations, and capture the dynamic evolution of moral judgments over time, which in turn will facilitate the Establishment, Evaluation, Embedding, Embodiment, Ensemble, and Evolvement (6Es) of the moral capabilities of AI models. Currently, STREAM has already furnished a comprehensive collection of ethical scenarios, and amassed substantial moral judgment data annotated by volunteers and various popular Large Language Models, collectively portraying the moral preferences and performances of both humans and AIs across a range of moral contexts. This paper will outline the current structure and construction of STREAM, explore its potential applications, and discuss its future prospects. (shrink)

Tennis players have more physical training content, and the training items are complex. For athletes, training programs that adapt to their individual characteristics should be formulated according to their physical characteristics. The current development of big data has brought about changes in thinking, management, and business models. The combination of complex systems and big data can also make breakthroughs in the sports field. Based on this, this article proposes a tennis player training schedule intelligent (...) formulation system based on complex system big data. First of all, this article adopts the literature data method, comparative analysis method, experimental analysis method, etc., in-depth study of the concepts of big data, complex system, and the physical structure characteristics of tennis sports. This paper designs an intelligent system for making tennis players’ training schedule, which collects, transforms, and integrates tennis training data through the characteristics of big data. Then, the dynamic time regulation of tennis is performed through a complex system, and finally, the experimental system is analyzed. This article mainly analyzes the comparison of physical indicators between the experimental group and the control group before and after the experiment, the evaluation indicators of sports events, the strength training effects of tennis events, and the analysis of shoulder joint tests. There is no significant difference between the experimental group and the control group in the items before the experiment, P > 0.05, which suggests that the physical fitness of the two groups of athletes is similar; in the posttest data, the experimental group and the control group have significant differences, P < 0.05, indicating an effect from the experiment. In particular, in fan running, forward and backward strokes, and the serve, the scores of the experimental group were higher than those of the control group, indicating that the use of the formulated training system demonstrated significant results. (shrink)

Nowadays, innovation and entrepreneurship courses occupy a very important place in universities and colleges and have also become an important teaching position in the process of building a new science. Colleges and universities actively respond to the challenge of “mass entrepreneurship and innovation” and define the goals and specifications of the talent training mechanism based on data fusion algorithms to cultivate as much high-quality applied talent as possible. In view of some shortcomings and problems in the current talent (...) training mechanism in universities and colleges, this paper proposes a data fusion algorithm based on information fusion theory and proof theory. The aim is to verify the feasibility of establishing a talent training mechanism for innovation and entrepreneurship education in universities and colleges. And this paper analyzes and explores the data fusion algorithm and the elements of innovation and entrepreneurial talent training, and forms an operating mechanism for entrepreneurial talent training according to social needs. Among them, the efficiency of the data fusion algorithm used by the GM model plays a significant role in the final result, and the minimum relative error value is 3.2%. Finally, it is concluded that we should focus on establishing a perfect talent training system for college students’ innovation and entrepreneurship education to improve students’ own comprehensive quality and various abilities, and to solve some social problems that are difficult to find employment in essence. (shrink)

Aims To examine the understanding of the ethical dilemmas associated with Big Data and artificial intelligence (AI) among Jordanian medical students, physicians in training, and senior practitioners. Methods We implemented a literature-validated questionnaire to examine the knowledge, attitudes, and practices of the target population during the period between April and August 2023. Themes of ethical debate included privacy breaches, consent, ownership, augmented biases, epistemology, and accountability. Participants’ responses were showcased using descriptive statistics and compared between groups using t-test (...) or ANOVA. Results We included 466 participants. The greater majority of respondents were interns and residents (50.2%), followed by medical students (38.0%). Most participants were affiliated with university institutions (62.4%). In terms of privacy, participants acknowledged that Big Data and AI were susceptible to privacy breaches (39.3%); however, 59.0% found such breaches justifiable under certain conditions. For ethical debacles involving informed consent, 41.6% and 44.6% were aware that obtaining informed consent posed an ethical limitation in Big Data and AI applications and denounced the concept of “broad consent”, respectively. In terms of ownership, 49.6% acknowledged that data cannot be owned yet accepted that institutions could hold a quasi-control of such data (59.0%). Less than 50% of participants were aware of Big Data and AI’s abilities to augment or create new biases in healthcare. Furthermore, participants agreed that researchers, institutions, and legislative bodies were responsible for ensuring the ethical implementation of Big Data and AI. Finally, while demonstrating limited experience with using such technology, participants generally had positive views of the role of Big Data and AI in complementing healthcare. Conclusion Jordanian medical students, physicians in training and senior practitioners have limited awareness of the ethical risks associated with Big Data and AI. Institutions are responsible for raising awareness, especially with the upsurge of such technology. (shrink)

The use of Big Data—however the term is defined—involves a wide array of issues and stakeholders, thereby increasing numbers of complex decisions around issues including data acquisition, use, and sharing. Big Data is becoming a significant component of practice in an ever-increasing range of disciplines; however, since it is not a coherent “discipline” itself, specific codes of conduct for Big Data users and researchers do not exist. While many institutions have created, or will create, training (...) opportunities to prepare people to work in and around Big Data, insufficient time, space, and thought have been dedicated to training these people to engage with the ethical, legal, and social issues in this new domain. Since Big Data practitioners come from, and work in, diverse contexts, neither a relevant professional code of conduct nor specific formal ethics training are likely to be readily available. This normative paper describes an approach to conceptualizing ethical reasoning and integrating it into training for Big Data use and research. Our approach is based on a published framework that emphasizes ethical reasoning rather than topical knowledge. We describe the formation of professional community norms from two key disciplines that contribute to the emergent field of Big Data: computer science and statistics. Historical analogies from these professions suggest strategies for introducing trainees and orienting practitioners both to ethical reasoning and to a code of professional conduct itself. We include two semester course syllabi to strengthen our thesis that codes of conduct can be harnessed to support the development of ethical reasoning in, and a sense of professional identity among, Big Data practitioners. (shrink)

In this chapter, we propose a non-traditional RCR training in data science that is grounded in a virtue theory framework. First, we delineate the approach in more theoretical detail by discussing how the goal of RCR training is to foster the cultivation of certain moral abilities. We specify the nature of these ‘abilities’: while the ideal is the cultivation of virtues, the limited space allowed by RCR modules can only facilitate the cultivation of superficial abilities or proto-virtues, (...) which help students to familiarize themselves with moral and political issues in the data science environment. Third, we operationalize our approach by stressing that (proto-)virtue acquisition (like skill acquisition) occurs through the technical and social tasks of daily data science activities, where these repetitive tasks provide the opportunities to develop (proto-)virtue capacity and to support the development of ethically robust data systems. Finally, we discuss a concrete example of implementing this approach. In particular, we describe how this method is applied to teach data ethics to students participating in the CODATA-RDA Data Science Summer Schools. (shrink)

In this study, we accessed information about the university credits of all teachers born after 1971 in Sweden as a means of ascertaining the development of their subject knowledge. We examined the university credits they earned during pre-service and in-service training. Data comes from registers Gothenburg Educational Longitudinal Database and the teacher register. We linked GOLD to the teacher register in order to describe the knowledge development of teachers in compulsory school 1998–2014. Special focus was on Swedish language (...) and mathematics. Multiple regression and multilevel growth modeling were used as our main methods. Results show an increase in pre-service credits during the time period and more credits in Swedish language than in mathematics. To analyze teachers’ in-service training, we followed the development of their university credits over time. Teachers with higher prerequisites in terms of grade point average tended to gain more credits in-service. The study included discussions on ideas and the implications for future research. (shrink)

Background: As actors with the key responsibility for the protection of human research participants, Research Ethics Committees (RECs) need to be competent and well-resourced in order to fulfil their roles. Despite recent programs designed to strengthen RECs in Africa, much more needs to be accomplished before these committees can function optimally.Objective: To assess training needs for biomedical research ethics evaluation among targeted countries.Methods: Members of RECs operating in three targeted African countries were surveyed between August and November 2007. Before (...) implementing the survey, ethical approvals were obtained from RECs in Switzerland, Cameroon, Mali and Tanzania. Data were collected using a semi-structured questionnaire in English and in French.Results: A total of 74 respondents participated in the study. The participation rate was 68%. Seventy one percent of respondents reported having received some training in research ethics evaluation. This training was given by national institutions (31%) and international institutions (69%). Researchers and REC members were ranked as the top target audiences to be trained. Of 32 topics, the top five training priorities were: basic ethical principles, coverage of applicable laws and regulations, how to conduct ethics review, evaluating informed consent processes and the role of the REC.Conclusion: Although the majority of REC members in the targeted African countries had received training in ethics, they expressed a need for additional training. The results of this survey have been used to design a training program in research ethics evaluation that meets this need. (shrink)

In this research, patent prosecution is conceptualized as a system of reinforcement learning from human feedback. The objective of the system is to increase the likelihood for a language model to generate patent claims that have a higher chance of being granted. To showcase the controllability of the language model, the system learns from granted patents and pre-grant applications with different rewards. The status of “granted” and “pre-grant” are perceived as labeled human feedback implicitly. In addition, specific to patent drafting, (...) the experiments in this research demonstrate the model’s capability to learn from adjusting claim length and inclusion of limiting terms for narrowing claim scope. As proof of concept, the experiments focus on claim ones only and the training data originates from a patent dataset tailored specifically for artificial intelligence. Although the available human feedback in patent prosecution are limited and the quality of generated patent text requires improvement, the experiments following the 3-stage reinforcement learning from human feedback have demonstrated that generative language models are capable of reflecting the human feedback or intent in patent prosecution. To enhance the usability of language models, the implementation in this research utilizes modern techniques that enable execution on a single consumer-grade GPU. The demonstrated proof of concept, which reduces hardware requirements, will prove valuable in the future as more human feedback in patent prosecution become available for broader use, either within patent offices or in the public domain. (shrink)

The pandemic caused by the COVID 19 Virus creates an unprecedented situation of global confinement altering the development of competition and sports training at all levels of participation and in all sports, including rhythmic gymnastics (RG). To avoid possible effects of physical, technical and psychological detraining, coaches looked for home training alternatives. The objectives of the study were to know how rhythmic gymnastics training developed during the lockdown period (the conditions, type of training, performance monitoring means, (...) and determinants of gymnasts’ participation) and to provide recommendations for a possible future lockdown. Three hundred and two RG coaches from twenty-six different countries throughout the five continents and four professional levels took part in the study: national team (28), international (26), national (172) and regional (75). The data collection tool was a questionnaire consisting of 39 closed questions structured in three dimensions: identification data of the coaches, training data during confinement and gymnast participation data. The independent variable was the gymnasts’ performance levels and the dependent variables organized in four categories: the technical media used to conduct and monitor the training sessions, the type of training done, the mechanisms for monitoring training performance and the aspects that determined participation. Most coaches kept their gymnasts training during confinement, although 76.5% confirm abandonment of any of their gymnasts. The main means used were real-time video conferencing, although at the lower practice levels the media stand out in deferred time. The contents of the training were mainly body technique, physical preparation and body difficulties. For performance monitoring, challenges, physical, and technical tests were predominant. The determinants for the development of training in the confinement vary depending on the level of the gymnasts, connectivity and electronic resources at the highest level, and the availability of spaces and social distancing at lower levels. For future lockdowns, it is necessary to review the content of the trainings, as well as the performance evaluation and the means necessary for it. (shrink)

This book addresses the challenges typical technical training managers, and other technical managers face in justifying the return on investment of their programs, particularly for large-scale, investment-intensive programs. This book describes a very intuitive and practical model for the measurement of the effectiveness of technical training programs. The book is based on a 4-tier Return on Effectiveness (ROE) model developed through years of research, observation, and experience. The ROE model uses four simple indices: training reaction index, improvement (...) index, effectiveness index, and impact trends to convey a full picture of the effectiveness of a training program. This feedback-based and data-based approach enables training managers with a step-by-step approach to collect data, measure important indicators and compute the indices to present evidence of the effectiveness of their training programs to the executives. (shrink)

This paper presents an ethical framework designed to support the development of critical data literacy for research methods courses and data training programmes in higher education. The framework we present draws upon our reviews of literature, course syllabi and existing frameworks on data ethics. For this research we reviewed 250 research methods syllabi from across the disciplines, as well as 80 syllabi from data science programmes to understand how or if data ethics was taught. (...) We also reviewed 12 data ethics frameworks drawn from different sectors. Finally, we reviewed an extensive and diverse body of literature about data practices, research ethics, data ethics and critical data literacy, in order to develop a transversal model that can be adopted across higher education. To promote and support ethical approaches to the collection and use of data, ethics training must go beyond securing informed consent to enable a critical understanding of the techno-centric environment and the intersecting hierarchies of power embedded in technology and data. By fostering ethics as a method, educators can enable research that protects vulnerable groups and empower communities. (shrink)

Parts I through III of this paper will examine several, increasingly comprehensive forms of aggregation, ranging from insurance reimbursement “lock-in” programs to PDMPs to completely unified electronic medical records. Each part will advocate for the adoption of these aggregation systems and provide suggestions for effective implementation in the fight against opioid misuse. All PDMPs are not made equal, however, and Part II will, therefore, focus on several elements — mandating prescriber usage, streamlining the user interface, ensuring timely data uploads, (...) creating a national data repository, mitigating privacy concerns, and training doctors on how to respond to perceived doctor-shopping — that can make these systems more effective. In each part, we will also discuss the privacy concerns of aggregating data, ranging from minimal to significant, and highlight the unique role of stigma in motivating these concerns. In Part IV, we will conclude by suggesting remedial steps to offset this loss of privacy and to combat the stigma around SUDs and mental health disorders in general. (shrink)

IntroductionThis special issue brings together researchers from psychology and linguistics who apply the ethnomethodologically informed analytic technique of conversation analysis (henceforth CA) to examine a range of ethical issues as they emerge in transcribed recordings of interactions collected as part of routine research encounters. The data authors analyse are diverse, including naturalistic audio and video recordings of members’ everyday and professional practices (Mondada 2014), an ethnography of a gynaecology unit in a public hospital in Italy (Fatigante and Orletti 2014), (...) focus group interviews on domestic family life (Alby and Fatigante 2014), and semi-public meetings, recorded as part of an ethnography of ageing, poverty, and social exclusion (Paoletti 2014b). Noting that “ethical issues in research have seldom been the focus of ethnomethodological studies” (Paoletti 2014a: 174; see also Mondada 2014: 182–183), the primary aim of this special issue is to “open a debate on e. (shrink)

There is a deluge of AI-assisted decision-making systems, where our data serve as proxy to our actions, suggested by AI. The closer we investigate our data (raw input, or their learned representations, or the suggested actions), we begin to discover “bugs”. Outside of their test, controlled environments, AI systems may encounter situations investigated primarily by those in other disciplines, but experts in those fields are typically excluded from the design process and are only invited to attest to the (...) ethical features of the resulting system or to comment on demonstrations of intelligence and aspects of craftmanship after the fact. This communicative impasse must be overcome. Our idea is that philosophical and engineering considerations interact and can be fruitfully combined in the AI design process from the very beginning. We embody this idea in the role of a philosopher engineer. We discuss the role of philosopher engineers in the three main design stages of an AI system: deployment management (what is the system’s intended use, in what environment?); objective setting (what should the system be trained to do, and how?); and training (what model should be used, and why?). We then exemplify the need for philosopher engineers with an illustrative example, investigating how the future decisions of an AI-based hiring system can be fairer than those contained in the biased input data on which it is trained; and we briefly sketch the kind of interdisciplinary education that we envision will help to bring about better AI. (shrink)

While often presented as promising assistive technologies for motor-impaired users, electroencephalography (EEG)-based Brain-Computer Interfaces (BCIs) remain barely used outside laboratories due to low reliability in real-life conditions. There is thus a need to design long-term reliable BCIs that can be used outside-of-the-lab by end-users, e.g., severely motor-impaired ones. Therefore, we propose and evaluate the design of a multi-class Mental Task (MT)-based BCI for longitudinal training (20 sessions over 3 months) of a tetraplegic user for the CYBATHLON BCI series 2019. (...) In this BCI championship, tetraplegic pilots are mentally driving a virtual car in a racing video game. We aimed at combining a progressive user MT-BCI training with a newly designed machine learning pipeline based on adaptive Riemannian classifiers shown to be promising for real-life applications. We followed a two step training process: the first 11 sessions served to train the user to control a 2-class MT-BCI by performing either two cognitive tasks (REST and MENTAL SUBTRACTION) or two motor-imagery tasks (LEFT-HAND and RIGHT-HAND). The second training step (9 remaining sessions) applied an adaptive, session-independent Riemannian classifier that combined all 4 MT classes used before. Moreover, as our Riemannian classifier was incrementally updated in an unsupervised way it would capture both within and between-session non-stationarity. Experimental evidences confirm the effectiveness of this approach. Namely, the classification accuracy improved by about 30% at the end of the training compared to initial sessions. We also studied the neural correlates of this performance improvement. Using a newly proposed BCI user learning metric, we could show our user learned to improve his BCI control by producing EEG signals matching increasingly more the BCI classifier training data distribution, rather than by improving his EEG class discrimination. However, the resulting improvement was effective only on synchronous (cue-based) BCI and it did not translate into improved CYBATHLON BCI game performances. For the sake of overcoming this in the future, we unveil possible reasons for these limited gaming performances and identify a number of promising future research directions. Importantly, we also report on the evolution of the user's neurophysiological patterns and user experience throughout the BCI training and competition. (shrink)

As the labeling costs in legal artificial intelligence tasks are expensive. Therefore, it becomes a challenge to utilize low cost to train a robust model. In this paper, we propose a LAIAugment approach, which aims to enhance the few-shot learning capability in legal artificial intelligence tasks. Specifically, we first use the self-training approach to label the amount of unlabelled data to enhance the feature learning capability of the model. Moreover, we also search for datasets that are similar to (...) the training set by improving the text similarity function. We conducted experimental analyses for three legal artificial intelligence tasks, including evidence extraction, legal element extraction, and case multi-label prediction, which composed of 3500 judgement documents. The experimental results show that the proposed LAIAugment method has an average F1-score of 72.3% on the three legal AI tasks, which is 1.93% higher than the baseline model. At the same time, it shows a huge improvement in few-shot learning. (shrink)

Changes in knowledge production, increasing interaction between government, universities and industry, and changes in labor markets for doctoral degree holders are forces that have spurred a debate about the organization of doctoral education and the competencies graduates need to master to work as scientists and researchers in a triple helix research context. Recent policy also has supported a redefinition of researcher training with increasing focus on broader skills and relevance for careers outside the university sector. Consequently, it is pertinent (...) to investigate current changes in doctoral education and researcher training. Particularly further knowledge about university–industry collaboration as a context for researcher training is required. With this in mind, this study provides empirical illustrations of how research training carried out in collaborative research contexts is experienced by doctoral students, and offers some insight into antecedent and process factors that are central in shaping PhD students’ research and training experience in collaborative research contexts. Based on the empirical data and a review of existing literature, suggestions for further research are made. (shrink)

Moral case deliberation is a dialogue among health care professionals about moral issues in practice. A trained facilitator moderates the dialogue, using a conversation method. Often, the facilitator is an ethicist. However, because of the growing interest in MCD and the need to connect MCD to practice, healthcare professionals should also become facilitators themselves. In order to transfer the facilitating expertise to health care professionals, a training program has been developed. This program enables professionals in health care institutions to (...) acquire expertise in dealing with moral questions independent of the expertise of an ethicist. Over the past 10 years, we developed a training program with a specific mix of theory and practice, aiming to foster the right attitude, skills and knowledge of the trainee. The content and the didactics of the training developed in line with the philosophy of MCD: pragmatic hermeneutics, dialogical ethics and Socratic epistemology. Central principles are: ‘learning by doing’, ‘reflection instead of ready made knowledge’, and ‘dialogue on dialogue’. This paper describes the theoretical background and the didactic content of the current training. Furthermore, we present didactic tools which we developed for stimulating active learning. We also go into lessons we learned in developing the training. Next, we provide some preliminary data from evaluation research of the training program by participants. The discussion highlights crucial aspects of educating professionals to become facilitators of MCD. The paper ends with concluding remarks and a plea for more evaluative evidence of the effectiveness and meaning of this training program for doing MCD in institutions. (shrink)

IntroductionClinical ethics committees (CECs) support and enhance communication and complex decision making, educate healthcare professionals and the public on ethical matters and maintain standards of care. However, a consistent approach to training members of CECs is lacking. A systematic scoping review was conducted to evaluate prevailing CEC training curricula to guide the design of an evidence-based approach.MethodsArksey and O’Malley’s methodological framework for conducting scoping reviews was used to evaluate prevailing accounts of CEC training published in six databases. (...) Braun and Clarke’s thematic analysis approach was adopted to thematically analyse data across different healthcare and educational settings.Results7370 abstracts were identified, 92 full-text articles were reviewed and 55 articles were thematically analysed to reveal four themes: the design, pedagogy, content and assessment of CEC curricula.ConclusionFew curricula employ consistent approaches to training. Many programmes fail to provide CEC trainees with sufficient knowledge, skills and experience to meet required competencies. Most programmes do not inculcate prevailing sociocultural, research, clinical and educational considerations into training processes nor provide longitudinal support for CEC trainees. Most CEC training programmes are not supported by host institutions threatening the sustainability of the programme and compromising effective assessment and longitudinal support of CEC trainees. While further reviews are required, this review underlines the need for host organisations to support and oversee a socioculturally appropriate ethically sensitive, clinically relevant longitudinal training, assessment and support process for CEC trainees if CECs are to meet their roles effectively. (shrink)

The present study examined the effects of mindfulness training on attention regulation in university students and whether the potential benefits of implementation are influenced by the yoga component of mindfulness-based interventions (MBIs) and/or by MBI homework practice. In a non-randomized trial with pre- and post-assessments, n = 180 university students were allocated to either mindfulness training (experimental groups), awareness activities (active control group), or no training (passive control group). Mindfulness was taught through two MBIs, one including yoga (...) and the other excluding yoga. Attention regulation was operationalized via behavioral indicators, namely sustained attention, cognitive flexibility, cognitive inhibition, and data-driven information processing. With the exception of speed in a cognitive flexibility task, the results indicated no systematic or differential advantage arising from mindfulness training, with or without yoga, regarding the aspects of attention regulation. There was no consistent influence of homework quantity or quality. The implications for mindfulness training in academic contexts are discussed. (shrink)

Ethics training has been highlighted as essential for building and fostering business ethics in organizations. National and international trends show that over 40% of businesses have some form of business ethics training. We use data collected from 199 firms to examine the presence of ethics training in top Canadian companies and found that the presence varied by region and firm size, and that the Canadian average (35%) lags other countries.

Historically, scientists in training have learned the rules of ethical conduct by the example of their advisors and other senior scientists and by practice. This paper is intended to serve as a guide for the beginning scientist to some fundamental principles of scientific research ethics. The paper focuses less on issues of outright dishonesty or fraud, and more on the positive aspects of ethical scientific behavior; in other words, what a scientist should do to maintain a high level of (...) ethical conduct in research.There are a number of fairly specific rules, guidelines, or commonly accepted operating principles that have evolved for the ethical conduct of science. In order to discuss this code of ethics, this paper is divided into sections dealing with specific areas of scientific ethics. These areas are: data collection and storage, ownership of data, confidentiality, communication, authorship, collaboration, the peer review system, and rules of dealing with ethical complaints. Illustrative case histories are presented to provide examples of the type of ethical dispute or problem being discussed. If scientific trainees learn the accepted rules of behavior that govern the conduct of science, ethical problems that arise out of ignorance, misunderstanding, or poor communication can be avoided. (shrink)

The advancement in technology has changed the workflow and the role of human translator in recent years. The impact from the trend of technology-mediated translation prompted the ratification of technology literacy as a major competence for modern translators. Consequently, teaching of translation technology including but not limited to Computer-aided Translation and Machine Translation became part of comprehensive curricula for translation training programs. However, in many institutions, the teaching of translation technology was haunted by issues such as: narrow scope of (...) curriculum design, outdated technologies, and unbalance between theories and practices in teaching. The study was the pilot evaluation of a tailored course to foster translation trainees’ knowledge and abilities of data science. The course was designed to be a fundamental step toward sophisticated translation technologies. During the pilot evaluation of the 8-week course, 85 students were recruited as participants. The study adopted a mix-method design by employing a survey to investigate student’s level of satisfaction toward the course and focus group discussion to understand students’ attitudes and perceptions of key aspects of the course. By interpreting the results from statistical analysis of the survey and thematic analysis of the focus group discussion, the course of data science for translators was well received among participants. The evaluation project manifested the feasibility and effectiveness of a translator-oriented data science course. (shrink)

El estudio persigue el objetivo de contribuir a la capacitación para elevar la competencia de los trabajadores del hospital general de Kuanza Sul, Angola, en la educación ambiental para la gestión de residuos sólidos en consonancia con la preservación del medio ambiente y la promoción de la salud de los pueblos. Se caracteriza al Hospital en sus dimensiones estructural y funcional. Contiene el análisis, la interpretación y el tratamiento de los datos obtenidos a través de encuestas, entrevistas y observación. This (...) study has the objective to contribute to train the workers at Kuanza Sul, General Hospital in environmental education for the management of solid waste in keeping with the protection of the environment and peoples health promotion. The hospital is characterized in its structural and functional dimensions. it contains the analysis, the interpretation and treatment of the data obtained by means of surveys, interviews and observation. (shrink)

Research on Augmented Reality in education has demonstrated that AR applications designed with diverse components boost student motivation in educational settings. However, most of the research conducted to date, does not define exactly what those components are and how these components positively affect student motivation. This study, therefore, attempts to identify some of the components that positively affect student motivation in mobile AR learning experiences to contribute to the design and development of motivational AR learning experiences for the Vocational Education (...) and Training level of education. To identify these components, a research model constructed from the literature was empirically validated with data obtained from two sources: 35 students from four VET institutes interacting with an AR application for learning for a period of 20 days, and a self-report measure obtained from the Instructional Materials Motivation Survey. We found that the following variables: use of scaffolding, real-time feedback, degree of success, time on-task and learning outcomes are positively correlated with the four dimensions of the ARCS model of motivation: Attention, Relevance, Confidence, and Satisfaction. Implications of these results are also described. (shrink)

Audiovisual communication is greatly contributing to the emerging research field of affective computing. The use of audiovisual stimuli within immersive virtual reality environments is providing very intense emotional reactions, which provoke spontaneous physical and physiological changes that can be assimilated into real responses. In order to ensure high-quality recognition, the artificial intelligence system must be trained with adequate data sets, including not only those gathered by smart sensors but also the tags related to the elicited emotion. Currently, there are (...) very few techniques available for the labeling of emotions. Among them, the Self-Assessment Manikin devised by Lang is one of the most popular. This study shows experimentally that the graphic proposal for the original SAM labelling system, as devised by Lang, is not neutral to gender and contains gender biases in its design and representation. Therefore, a new graphic design has been proposed and tested according to the guidelines of expert judges. The results of the experiment show an overall improvement in the labeling of emotions in the pleasure–arousal–dominance affective space, particularly, for women. This research proves the relevance of applying the gender perspective in the validation of tools used throughout the years. (shrink)

The quality and quantity of training data are crucial to the performance of a deep-learning-based brain-computer interface system. However, it is not practical to record EEG data over several long calibration sessions. A promising time- and cost-efficient solution is artificial data generation or data augmentation. Here, we proposed a DA method for the motor imagery EEG signal called brain-area-recombination. For the BAR, each sample was first separated into two ones by left/right brain channels, and the (...) artificial samples were generated by recombining the half-samples. We then designed two schemas corresponding to the single- and multi-subject scenarios. Extensive experiments using the classifier of EEGnet were conducted on two public datasets under various training set sizes. In both schemas, the BAR method can make the EEGnet have a better performance of classification. To make a comparative investigation, we selected two common DA methods, and the BAR method beat them. Further, using the proposed BAR for augmentation, EEGnet achieved up to 8.3% improvement than a typical decoding algorithm CSP-SVM, note that both the models were trained on the augmented dataset. This study shows that BAR usage can significantly improve the classification ability of deep learning to MI-EEG signals. To a certain extent, it may promote the development of deep learning technology in the field of BCI. (shrink)

Public sector adoption of AI techniques in welfare systems recasts historic national data as resource for machine learning. In this paper, we examine how the use of register data for development of predictive models produces new ‘afterlives’ for citizen data. First, we document a Danish research project’s practical efforts to develop an algorithmic decision-support model for social workers to classify children’s risk of maltreatment. Second, we outline the tensions emerging from project members’ negotiations about which datasets to (...) include. Third, we identify three types of afterlives for citizen data in machine learning projects: (1) data afterlives for training and testing the algorithm, acting as ‘ground truth’ for inferring futures, (2) data afterlives for validating the algorithmic model, acting as markers of robustness, and (3) data afterlives for improving the model’s fairness, valuated for reasons of data ethics. We conclude by discussing how, on one hand, these afterlives engender new ethical relations between state and citizens; and how they, on the other hand, also articulate an alternative view on the value of datasets, posing interesting contrasts between machine learning projects developed within the context of the Danish welfare state and mainstream corporate AI discourses of the bigger, the better. (shrink)

Impairments in executive functioning are prevalent in chronic pain conditions, with cognitive inflexibility being the most frequently reported. The current randomized, cross-over trial, piloted a computerized cognitive training program based on Relational Frame Theory, targeting improvement in cognitive flexibility. At baseline, 73 chronic pain patients completed testing on pre-selected outcomes of executive functioning, alongside IQ measures. When tested three times over the course of 5 months, there was a drop-out rate of 40% at the third time point, leaving 44 (...) patients who had data at all time points. The results showed that there was a substantial learning effect from the MINDFLEX training and a substantial time-dependent improvement on the primary outcomes of increased flexibility, but that this could not be tied to active training. In conclusion, this small study indicated a learning effect as well as improvement on primary outcomes. Based on the current results, a larger trial with improved feasibility of training is warranted. (shrink)

A model frequently used to implement community‐based research involves engaging local community health workers who are trusted members of the community and familiar with local customs, language, and culture. In Spanish‐speaking communities, the CHWs are also known as promotores de salud (“health promoters”). Depending on the study design and nature of the research, promotores facilitate research through community outreach, instrument design, participant recruitment, intervention delivery, data collection, and other research‐related tasks. In 2000, the National Institutes of Health published a (...) regulation requiring training of “key personnel” named on NIH‐supported research involving human subjects. Regardless of whether promotores are technically designated as key personnel on a particular study, they may benefit from training in research ethics. Unfortunately, the educational programs designed for academic researchers in response to the NIH mandate were poorly aligned with the needs of promotores and not suited for learning about human research ethics applied to the community health setting. For example, many promotores in our California‐Mexico border region are monolingual Spanish speakers who have no formal academic research training. Therefore, we set out to develop a training alternative appropriate to our local CHWs. In this essay, we provide overviews of the relevant instructional design principles and our formative research process, including examples of key findings. (shrink)

The article summarizes the theoretical and methodological knowledge about ecotourism as one of the viable types of tourism in the postmodern era, clarifies the patterns of its formation and development, a variety of approaches to its interpretation, interaction with other types of tourism, features of motivation and management in ecotourism. On the basis of the generalized data a number of perspective educational conditions is modeled: the orientation of the maintenance of pedagogical education on formation of steady positive motivation; updating (...) the content of professional training; availability of educational and methodical support; interrelation of competence, system and personal-activity approaches. Diagnostic methods relevant for postmodern education are selected: methods of diagnosing students' learning motivation; methods of diagnosis of value orientations in career; tests to determine the levels of formation of the components of professional competence. It is proved that the professional training of teachers who are able to use ecological tourism in their professional activities at a high level is one of the important conditions for increasing the interest of children and young people in physical education, tourism and sports, ensuring the necessary physical activity, developing, maintaining and strengthening their health, adaptation to the flexible postmodern society. (shrink)

By investigating the status quo of the swimming training market in a certain area, we can obtain information on the current development of the swimming training market in a certain area and study the laws of the development of the market so as to provide a theoretical basis for the development of the market. This paper designs an evaluation algorithm suitable for swimming training based on the improved AlexNet network. The algorithm model uses a 3 × 3 (...) size convolution kernel to extract features, and the pooling layer uses a nonoverlapping pooling strategy. In order to accelerate the network convergence, the model introduces batch normalization technology. The algorithm uses data augmentation technology to expand the data set, including rotation and random erasure, to a certain extent alleviating the problem of overfitting. The results of the study showed that there were no significant differences in fat, minerals, protein, body mass index, basal metabolic rate, and total energy expenditure in the body composition ratios of children in the convolutional neural network assessment group and the control group, while muscle and total body water were not significantly different. However, there are significant differences in fat-free body weight and muscle strength of various segments of the body, among which there are very significant differences in muscle strength of lower limbs in each segment of the body. There were no significant differences in minerals, body mass index, basal metabolic rate, total energy expenditure, and lower limb muscle strength in the body composition ratios of men and women in the convolutional neural network assessment group. There are significant differences in body weight, upper limb muscle strength, and trunk muscle strength. There were no significant differences in the proportions of body composition between men and women in the control group, except for fat and protein. (shrink)

Data sharing – broadly defined as the exchange of health-related data among multiple controllers and processors – has gained increased relevance in the health sciences over recent years as the need and demand for collaboration has increased. This includes data obtained through healthcare provisions, clinical trials, observational studies, public health surveillance programs, and other data collection methods. The practice of data sharing presents several notable challenges, however. Compliance with a complex and dynamic regulatory framework is (...) essential, with the General Data Protection Regulation being a prominent example in a European context. Recent regulatory developments related to clinical trial transparency, trade secrecy, data access, AI training data, and health data spaces further contribute to the difficulties. Simultaneously, government initiatives often encourage scientists to embrace principles of “open data” and “open innovation.” The variety of regulations in this domain has the potential to impede widespread data sharing and hinder innovation. This edited volume, therefore, compiles comparative case studies authored by leading scholars from diverse disciplines and jurisdictions. The book aims to outline the legal complexities of data sharing. By examining real-world scenarios from diverse disciplines and a global perspective, it explores the normative, policy, and ethical dilemmas that surround data sharing in the health sciences today. Chapter Patient Perspectives on Data Sharing, Chapter Supplementary Measures and Appropriate Safeguards for International Transfers of Health Data after Schrems II are available open access under a Creative Commons Attribution 4.0 International License via link.springer.com. (shrink)

The advent of artificial intelligence (AI) challenges political theorists to think about data ownership and policymakers to regulate the collection and use of public data. AI producers benefit from free public data for training their systems while retaining the profits. We argue against the view that the use of public data must be free. The proponents of unconstrained use point out that consuming data does not diminish its quality and that information is in ample (...) supply. Therefore, they suggest, publicly available data should be free. We present two objections. First, allowing free data use promotes unwanted inequality. Second, contributors of information did not and could not anticipate that their contribution would be used to train AI systems. Therefore, charging for extensive data use is pro tanto permissible and desirable. We discuss policy implications and propose a progressive data use tax to counter the inequality arising. (shrink)

Denmark is a society that has already moved towards Big Data and a Learning Health Care System. Data from routine healthcare has been registered centrally for years, there is a nationwide tissue bank, and there are numerous other available registries about education, employment, housing, pollution, etcetera. This has allowed Danish researchers to study the link between exposures, genetics and diseases in a large population. This use of public registries for scientific research has been relatively uncontroversial and has been (...) supported by facilitative regulation that allows data to be used without the consent of data subjects. However, in the future much of the data will not be held by public authorities but by private companies. What are the implications of this shift for the governance of the research use of the data? This paper will argue that increased involvement of Research Ethics Committees and better training of researchers are necessary; and that some form of consent will have to be re-introduced. Four different consent models will be discussed: Opt-Out, Broad/Blanket consent, Dynamic consent, and Meta consent. It will be argued that a governance model including a possibility for citizens to make meta-choices strikes the best balance between individual and public interests. (shrink)

According to the expertise defense, practitioners of the method of cases need not worry about findings that ordinary people’s philosophical intuitions depend on epistemically irrelevant factors. This is because, honed by years of training, the intuitions of professional philosophers likely surpass those of the folk. To investigate this, we conducted a controlled longitudinal study of a broad range of intuitions in undergraduate students of philosophy (n = 226), whose case judgments we sampled after each semester throughout their studies. Under (...) the assumption, made by proponents of the expertise defense, that formal training in philosophy gives rise to the kind of expertise that accounts for changes in the students’ responses to philosophically puzzling cases, our data suggest that the acquired cognitive skills only affect single case judgments at a time. There does not seem to exist either a general expertise that informs case judgments in all areas of philosophy, or an expertise specific to particular subfields. In fact, we argue that available evidence, including the results of cross-sectional research, is best explained in terms of differences in adopted beliefs about specific cases, rather than acquired cognitive skills. We also investigated whether individuals who choose to study philosophy have atypical intuitions compared to the general population and whether students whose intuitions are at odds with textbook consensus are more likely than others to drop out of the philosophy program. (shrink)