30115 research outputs found
Sort by
A digital companion for musicological scholarship: the Lohengrin TimeMachine
Get PDFMusicology has long suffered from the difficulties of making its work accessible, or even comprehensible to a wider audience. We introduce a digital companion to music scholarship – in this case, an exploration of Wagner's early use of leading motifs in the opera Lohengrin – providing different ways to explore, see and hear the materials that are discussed in musicological research. That research is described in text and a video, each of which is incorporated into the companion as a springboard for further discovery. Novel visualisations include a ‘TimeMachine’ view, in which a user can flick through motif occurrences, quickly navigating through the musical transformations across the opera. Having introduced the scholarship and the companion, we discuss the collaborative process by which the application was conceived and built, including the practicalities of timing and incorporating external design expertise. We conclude by discussing the future of such ‘companions’ in musicological publication
Scientific hypothesis generation by large language models: laboratory validation in breast cancer treatment
Get PDFLarge language models (LLMs) have transformed artificial intelligence (AI) and achieved breakthrough performance on a wide range of tasks. In science, the most interesting application of LLMs is for hypothesis formation. A feature of LLMs, which results from their probabilistic structure, is that the output text is not necessarily a valid inference from the training text. These are termed ‘hallucinations’, and are harmful in many applications. In science, some hallucinations may be useful: novel hypotheses whose validity may be tested by laboratory experiments. Here, we experimentally test the application of LLMs as a source of scientific hypotheses using the domain of breast cancer treatment. We applied the LLM GPT4 to hypothesize novel synergistic pairs of US Food and Drug Administration (FDA)-approved non-cancer drugs that target the MCF7 breast cancer cell line relative to the non-tumorigenic breast cell line MCF10A. In the first round of laboratory experiments, GPT4 succeeded in discovering three drug combinations (out of 12 tested) with synergy scores above the positive controls. GPT4 then generated new combinations based on its initial results, this generated three more combinations with positive synergy scores (out of four tested). We conclude that LLMs are a valuable source of scientific hypotheses
Introduction to NLP in Finance: Sentiment Analysis and Risk Management
No full textFinancial knowledge and understanding have become increasingly important in today's complex world. The ability to manage money effectively is no longer just a personal skill; it's essential for navigating the challenges and opportunities of modern life, but not everyone has the ability to understand and manage personal finances effectively. Financial market has undergone different ages, with the advancement in technologies, many automated tools can help even the financial il-literate people to understand the market dynamics. The advent of social media has revolutionized numerous aspects of human life, from communication to commerce. One of the most profound impacts has been on the dissemination and consumption of information, particularly in the realm of finance. To make a good investment decision, effective market surveillance systems are needed, that can analyze information related to a particular stock like financial documents, earning reports, news, market sentiment etc. Financial markets are driven by in-formation, much of which is textual data like news articles, social media posts, and company filings. Financial experts and enthusiasts can now share their knowledge and insights through blogging platforms, podcasts, and video content. This has made complex financial concepts more accessible to a wider audience, empowering individuals to make informed financial decisions. Natural language processing helps in sentiment analysis, risk management, and economic forecasting by analyzing those textual documents. This chapter presents different applications of natural language processing in financial analysis, along with the recent research work conducted in the field
Classifying Cognitive States of Alzheimer’s Disease with Machine Learning Using Digital Biomarkers from the Bio-Hermes Study Cohort
No full textAlzheimer’s Dementia (AD) is a progressive neurodegenerative disorder with clinical stages ranging from cognitively normal (CN) to mild cognitive impairment (MCI) and mild Alzheimer’s Dementia (AD). Accurate classification of these stages is critical for early intervention and treatment planning, and AD risk prediction is also important for patient stratification. This work explores the predictive capabilities of machine learning models using a combination of digital biomarkers from the Bio-Hermes study. Five models–Logistic Regression, Elastic Net, Random Forest, Gradient Boosting, and XGBoost–were evaluated on a 20% test set after being trained or hyperparameter-tuned using repeated cross-validation, on an 80% training set. XGBoost achieved the best overall performance with an AUC of 0.815 followed closely by Gradient Boosting (AUC = 0.806). The findings highlight the potential of ensemble methods like XGBoost for the prediction of Alzheimer’s clinical stages and risk of AD, with digital biomarkers proving to be valuable while non-invasive and inexpensive to produce predictors. This is an important aspect in the economy of a very expensive and challenging to diagnose, treat, and evaluate risk for, medical condition such as Dementia
Avatars in mixed-reality meetings: A longitudinal field study of realistic versus cartoon facial likeness effects on communication, task satisfaction, presence, and emotional perception
Get PDFWe conducted a within-subjects study to examine how realistic faces and cartoon faces on avatars affect communication, task satisfaction, sense of presence, and mood perception in mixed reality meetings. Over the course of two weeks, six groups of co-workers (14 people) held recurring meetings using Microsoft HoloLens2 devices, each person embodying a personal full-body avatar with either a realistic face or cartoon face. Half of the groups started with the realistic face avatar and switched to the cartoon face version halfway through (RC condition), and the other half with the cartoon-face avatar first (CR condition). Results showed that participants in the RC condition may have had higher expectations and more errors in perceiving their colleagues’ moods. Participants in the CR condition reported that the avatars’ appearance mattered less over time and experienced increased comfort and improved identification of their colleagues. Participants rated words, tone of voice, and movement as the most useful cues for perceiving colleagues’ moods, regardless of avatar rendering style. In the RC condition, participants rated gaze as more useful than facial expressions, while in the CR condition, both gaze and facial expressions were rated as the least useful. Results also suggested that participants had more errors when perceiving negative moods in their colleagues, with this trend appearing for most moods, but depending on conditions. Implications of these findings for mixed and virtual reality meetings are discussed. This work contributes to the field of remote collaboration by providing insights from longitudinal data on the impact of avatar appearance on various aspects of work meetings in virtual environments
Shaping the future: principles for policy recommendations for responsible innovation in virtual worlds
Get PDFExtended Reality (XR) technologies are beginning to enter the mainstream and have the potential to change the way humans interact with computers on a global scale. As with all powerful tools, XR not only has the potential for enormous good, but also brings in a new set of challenges for policy to guide its innovation and development. This paper draws on a wide range of expertise from academia, research and development, and industry to collectively provide guiding principles for XR policy. The authors began discussions and developed a framework at a workshop at ACM SIGCHI 2024 and the ideas presented here are the result of debate, discussion, refinement, and offer next steps for the development of pervasive XR. We present three main principles for XR policy: Trust, Agency, and Inclusivity, along with a cross-cutting theme of Future-Proofing. Each principle is broken down and we offer example implementations. This paper aims to build upon previous work and efforts for fair and equitable XR for all, and further dialogue towards tangible changes in policy to help guide responsible innovation in virtual worlds
Causal Inference in HR Analytics with Directed Acyclic Graphs
No full textPeople analytics is a nascent field that has yet to embrace the topic of causal inference. Yet understanding causality is often critically important in human resources. In some areas of management, accurate training predictions from a machine learning model might be enough to justify a decision, such as about a website design to maximize sales. As long as sales increase, it doesn’t really matter what causes customers to buy more on one website compared to the next. In contrast, with people analytics problems, causes do matter. Practitioners often look to change behavior with interventions, which requires an understanding of causality. Consider the results of an engagement survey, where engagement is found to be correlated with employee turnover. Is low engagement the cause of higher turnover, or is it merely associated with it, with other unobserved variables—like poor line management—perhaps causing both low engagement and high turnover? The answer determines the right action to reduce turnover
Tactical Learning and Innovation in the Habsburg Army, 1914-1918
No full textThe paper examines the wartime Habsburg Army as a learning organisation. It explores how the force drew tactical lessons from combat, from allies and from enemies and how it attempted to disseminate better practices and improve its combat performance during the First World War
‘E Pluribus Unum......Forum?’: A Marxist Approach to the EU’s Democratic Deficit
Get PDFThis paper approaches the issue of the European Union’s (EU) democratic deficit from a Marxist perspective. This issue has been central to the exponential rise of Euroscepticism that influenced processes like Brexit and Grexit (despite the latter’s frustration), as well as the rise of explicitly anti-EU national governments in European countries. This paper shows that critiques of the EU’s democratic deficit (even cutting-edge ones, like the one placing emphasis on the notion of the ‘economic constitution’) are inadequate because the debate is already embedded in ideological compromise. Offering a brief exposition of the Marxist approach to the democratic form of the capitalist state, it attempts to show the limitations of critical approaches which overlook the issue of class rule and state power in their calls for democratisation. To do so, the paper outlines the structural function and class character of the EU, as well as its role as a (supra-)state formation in the process of capital accumulation. Ultimately, it offers a Janus-faced critique of democratic deficit in Europe, one the one hand arguing that the critique of the EU economic constitution as neoliberal is limited because it fails to account for the scope of reform that the EU allows to respond to the challenges of the process of capital accumulation, while on the other concluding that the solution to the democratic deficit cannot simply be a return to nation-state democracy which is equidistant from actual self-government of the popular strata as its EU counterpart