Copilot in service: Exploring the potential of the large language model-based chatbots for fostering evaluation culture in preventing and countering violent extremism DOI Creative Commons

Irina van der Vet,

Leena Malkki

Open Research Europe, Год журнала: 2025, Номер 5, С. 65 - 65

Опубликована: Апрель 17, 2025

Background The rapid advancement in artificial intelligence (AI) technology has introduced the large language model (LLM)-based assistants or chatbots. To fully unlock potential of this for preventing and countering violent extremism (P/CVE) field, more research is needed. This paper examines feasibility using chatbots as recommender systems to respond practitioners’ needs evaluation, increase their knowledge about key evaluation aspects, provide practical guidance professional support process. At same time, provides an overview limitations that such solution entails. Methods explore performance LLM-based we chose a publicly available AI assistant called Copilot example. We conducted qualitative analysis its responses 50 pre-designed prompts various types. study was driven by questions established accuracy reliability, relevance integrity, well readability comprehensiveness responses. derived aspects evidence-based along with from results H2020 INDEED project. Results Our findings indicate demonstrated significant proficiency addressing issues related P/CVE. Most generated were factually accurate, relevant, structurally sound, i.e. sufficient kick-start deepen internal practise. biases data security inherent should be carefully explored practitioners. Conclusions underscored both fostering culture While can effectively generate accessible, informative encouraging recommendations, it still requires oversight manage coordinate process, address field-specific needs. future focus on rigorous user-centred assessment P/CVE use based multidisciplinary efforts.

Язык: Английский

Quantum Computing and Machine Learning in Medical Decision-Making: A Comprehensive Review DOI Creative Commons
James C. L. Chow

Algorithms, Год журнала: 2025, Номер 18(3), С. 156 - 156

Опубликована: Март 9, 2025

Medical decision-making is increasingly integrating quantum computing (QC) and machine learning (ML) to analyze complex datasets, improve diagnostics, enable personalized treatments. While QC holds the potential accelerate optimization, drug discovery, genomic analysis as hardware capabilities advance, current implementations remain limited compared classical in many practical applications. Meanwhile, ML has already demonstrated significant success medical imaging, predictive modeling, decision support. Their convergence, particularly through (QML), presents opportunities for future advancements processing high-dimensional healthcare data improving clinical outcomes. This review examines foundational concepts, key applications, challenges of these technologies healthcare, explores their synergy solving problems, outlines directions quantum-enhanced decision-making.

Язык: Английский

Процитировано

2
Copilot in service: Exploring the potential of the large language model-based chatbots for fostering evaluation culture in preventing and countering violent extremism DOI Creative Commons

Irina van der Vet,

Leena Malkki

Open Research Europe, Год журнала: 2025, Номер 5, С. 65 - 65

Опубликована: Апрель 17, 2025

Background The rapid advancement in artificial intelligence (AI) technology has introduced the large language model (LLM)-based assistants or chatbots. To fully unlock potential of this for preventing and countering violent extremism (P/CVE) field, more research is needed. This paper examines feasibility using chatbots as recommender systems to respond practitioners’ needs evaluation, increase their knowledge about key evaluation aspects, provide practical guidance professional support process. At same time, provides an overview limitations that such solution entails. Methods explore performance LLM-based we chose a publicly available AI assistant called Copilot example. We conducted qualitative analysis its responses 50 pre-designed prompts various types. study was driven by questions established accuracy reliability, relevance integrity, well readability comprehensiveness responses. derived aspects evidence-based along with from results H2020 INDEED project. Results Our findings indicate demonstrated significant proficiency addressing issues related P/CVE. Most generated were factually accurate, relevant, structurally sound, i.e. sufficient kick-start deepen internal practise. biases data security inherent should be carefully explored practitioners. Conclusions underscored both fostering culture While can effectively generate accessible, informative encouraging recommendations, it still requires oversight manage coordinate process, address field-specific needs. future focus on rigorous user-centred assessment P/CVE use based multidisciplinary efforts.

Язык: Английский

Процитировано

0