Context-Aware Offensive Language Detection in Human-Chatbot Conversations DOI
Mingi Shin, Hyojin Chin, Hyeonho Song

et al.

Published: Feb. 18, 2024

Dialogs generated by chatbots may contain unethical and offensive language that can negatively affect users, the service, society. Existing methods for automatically detecting are not effective chat data, which is short multi-turn hence requires understanding subtle context behind language. We introduce a new dataset from real human-chatbot conversations with context-aware annotations identify kinds of only in certain context. propose neural network model CALIOPER (Context-Aware modeL Identifying Offensive using Pre-trained Encoder Retrieval), uses encoder attention mechanism to incorporate previous messages retrieve relevant information implicit offensiveness. Experimental results show performs well on dialog par-ticularly context-dependent This work contributes making safer chatbot ecosystem advancing techniques detect data. (Disclaimer: contains profanity due study topic, we replace * marks.)

Language: Английский

Impact of ChatGPT on medical chatbots as a disruptive technology DOI Creative Commons
James C. L. Chow, Leslie Sanders, Kay Li

et al.

Frontiers in Artificial Intelligence, Journal Year: 2023, Volume and Issue: 6

Published: April 5, 2023

OPINION article Front. Artif. Intell., 05 April 2023Sec. Medicine and Public Health Volume 6 - 2023 | https://doi.org/10.3389/frai.2023.1166014

Language: Английский

Citations

128
The effects of artificial intelligence applications in educational settings: Challenges and strategies DOI Creative Commons
Omar Ali, Peter Murray, Mujtaba M. Momin

et al.

Technological Forecasting and Social Change, Journal Year: 2023, Volume and Issue: 199, P. 123076 - 123076

Published: Dec. 14, 2023

With the continuous intervention of AI tools in education sector, new research is required to evaluate viability and feasibility extant platforms inform various pedagogical methods instruction. The current manuscript explores cumulative published literature date order key challenges that influence implications adopting models Education Sector. researchers' present works both favour against AI-based applications within Academic milieu. A total 69 articles from a 618-article population was selected diverse academic journals between 2018 2023. After careful review articles, presents classification structure based on five distinct dimensions: user, operational, environmental, technological, ethical challenges. recommends use ChatGPT as complementary teaching-learning aid including need afford customized optimized versions tool for teaching fraternity. study addresses an important knowledge gap how enhance educational settings. For instance, discusses interalia range AI-related effects learning creative prompts, training datasets genres, incorporation human input data confidentiality elimination bias. concludes by recommending strategic solutions emerging identified while summarizing ways encourage wider adoption other sector. insights presented this can act reference policymakers, teachers, technology experts stakeholders, facilitate means sector more generally. Moreover, provides foundation future research.

Language: Английский

Citations

73
The Potential of Chatbots for Emotional Support and Promoting Mental Well-Being in Different Cultures: Mixed Methods Study DOI Creative Commons
Hyojin Chin, Hyeonho Song, Gumhee Baek

et al.

Journal of Medical Internet Research, Journal Year: 2023, Volume and Issue: 25, P. e51712 - e51712

Published: Sept. 29, 2023

Artificial intelligence chatbot research has focused on technical advances in natural language processing and validating the effectiveness of human-machine conversations specific settings. However, real-world chat data remain proprietary unexplored despite their growing popularity, new analyses uses effects mitigating negative moods are urgently needed.In this study, we investigated whether how artificial chatbots facilitate expression user emotions, specifically sadness depression. We also examined cultural differences depressive among users Western Eastern countries.This study used SimSimi, a global open-domain social chatbot, to analyze 152,783 conversation utterances containing terms "depress" "sad" 3 countries (Canada, United Kingdom, States) 5 (Indonesia, India, Malaysia, Philippines, Thailand). Study 1 reports findings people talk about depression based Linguistic Inquiry Word Count n-gram analyses. In 2, classified into predefined topics using semisupervised classification techniques better understand types prevalent chats. then identified distinguishing features chat-based discourse disparity between users.Our revealed intriguing differences. Chatbot indicated stronger emotions than (positive: P<.001; negative: P=.01); for example, more words associated with (P=.01). were likely share vulnerable such as mental health (P<.001), group had greater tendency discuss sensitive swear (P<.001) death (P<.001). addition, when talking chatbots, expressed differently other platforms. Users open expressing emotional vulnerability related or sad (74,045/148,590, 49.83%) media (149/1978, 7.53%). tended not broach that require support from others, seeking advice daily life difficulties, unlike media. acted anticipation conversational agents exhibit active listening skills foster safe space where they can openly states depression.The highlight potential chatbot-assisted support, emphasizing importance continued policy-wise efforts improve interactions those need assistance. Our indicate possibility providing helpful information moods, especially who have difficulty communicating humans.

Language: Английский

Citations

44
Roles, Users, Benefits, and Limitations of Chatbots in Health Care: Rapid Review DOI Creative Commons
Moustafa Laymouna, Yuanchao Ma, David Lessard

et al.

Journal of Medical Internet Research, Journal Year: 2024, Volume and Issue: 26, P. e56930 - e56930

Published: April 12, 2024

Background Chatbots, or conversational agents, have emerged as significant tools in health care, driven by advancements artificial intelligence and digital technology. These programs are designed to simulate human conversations, addressing various care needs. However, no comprehensive synthesis of chatbots’ roles, users, benefits, limitations is available inform future research application the field. Objective This review aims describe characteristics, focusing on their diverse roles pathway, user groups, limitations. Methods A rapid published literature from 2017 2023 was performed with a search strategy developed collaboration sciences librarian implemented MEDLINE Embase databases. Primary studies reporting chatbot benefits were included. Two reviewers dual-screened results. Extracted data subjected content analysis. Results The categorized into 2 themes: delivery remote services, including patient support, management, education, skills building, behavior promotion, provision administrative assistance providers. User groups spanned across patients chronic conditions well cancer; individuals focused lifestyle improvements; demographic such women, families, older adults. Professionals students also alongside seeking mental behavioral change, educational enhancement. chatbots classified improvement quality efficiency cost-effectiveness delivery. identified encompassed ethical challenges, medicolegal safety concerns, technical difficulties, experience issues, societal economic impacts. Conclusions Health offer wide spectrum applications, potentially impacting aspects care. While they promising for improving quality, integration system must be approached consideration ensure optimal, safe, equitable use.

Language: Английский

Citations

31
Generative Pre-Trained Transformer-Empowered Healthcare Conversations: Current Trends, Challenges, and Future Directions in Large Language Model-Enabled Medical Chatbots DOI Creative Commons
James C. L. Chow, Valerie Wong, Kay Li

et al.

BioMedInformatics, Journal Year: 2024, Volume and Issue: 4(1), P. 837 - 852

Published: March 14, 2024

This review explores the transformative integration of artificial intelligence (AI) and healthcare through conversational AI leveraging Natural Language Processing (NLP). Focusing on Large Models (LLMs), this paper navigates various sections, commencing with an overview AI’s significance in role AI. It delves into fundamental NLP techniques, emphasizing their facilitation seamless conversations. Examining evolution LLMs within frameworks, discusses key models used healthcare, exploring advantages implementation challenges. Practical applications conversations, from patient-centric utilities like diagnosis treatment suggestions to provider support systems, are detailed. Ethical legal considerations, including patient privacy, ethical implications, regulatory compliance, addressed. The concludes by spotlighting current challenges, envisaging future trends, highlighting potential reshaping interactions.

Language: Английский

Citations

27
Factuality challenges in the era of large language models and opportunities for fact-checking DOI
Isabelle Augenstein, Timothy Baldwin, Meeyoung Cha

et al.

Nature Machine Intelligence, Journal Year: 2024, Volume and Issue: 6(8), P. 852 - 863

Published: Aug. 22, 2024

Language: Английский

Citations

17
The knowledge and innovation challenges of ChatGPT: A scoping review DOI
Omar Ali, Peter Murray, Mujtaba M. Momin

et al.

Technology in Society, Journal Year: 2023, Volume and Issue: 75, P. 102402 - 102402

Published: Oct. 21, 2023

Language: Английский

Citations

37
The Chatbots Are Invading Us: A Map Point on the Evolution, Applications, Opportunities, and Emerging Problems in the Health Domain DOI Creative Commons
Daniele Giansanti

Life, Journal Year: 2023, Volume and Issue: 13(5), P. 1130 - 1130

Published: May 5, 2023

The inclusion of chatbots is potentially disruptive in society, introducing opportunities, but also important implications that need to be addressed on different domains. aim this study examine in-depth, by mapping out their technological evolution, current usage, and potential applications, emerging problems within the health domain. examined three points view. first point view traces evolution chatbots. second reports fields application chatbots, giving space expectations use expected benefits from a cross-domain view, affecting third main analysis state domain based scientific literature represented systematic reviews. overview identified topics greatest interest with opportunities. revealed for initiatives simultaneously evaluate multiple domains all together synergistic way. Concerted efforts achieve are recommended. It believed monitor both process osmosis between other sectors domain, as well can create psychological behavioural an impact

Language: Английский

Citations

23
Application of artificial intelligence (AI) to control COVID-19 pandemic: Current status and future prospects DOI Creative Commons
Sumel Ashique, Neeraj Mishra, Sourav Mohanto

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(4), P. e25754 - e25754

Published: Feb. 1, 2024

The impact of the coronavirus disease 2019 (COVID-19) pandemic on everyday livelihood people has been monumental and unparalleled. Although vastly affected global healthcare system, it also a platform to promote develop pioneering applications based autonomic artificial intelligence (AI) technology with therapeutic significance in combating pandemic. Artificial successfully demonstrated that can reduce probability human-to-human infectivity virus through evaluation, analysis, triangulation existing data spread virus. This review talks about modern robotic automated systems may assist spreading In addition, this study discusses intelligent wearable devices how they could be helpful throughout COVID-19

Language: Английский

Citations

9
Can ChatGPT replace humans in crisis communication? The effects of AI-mediated crisis communication on stakeholder satisfaction and responsibility attribution DOI Creative Commons
Yi Xiao, Shubin Yu

International Journal of Information Management, Journal Year: 2024, Volume and Issue: 80, P. 102835 - 102835

Published: Aug. 30, 2024

Imagine a world where chatbots are the first responders to crises, efficiently addressing concerns and providing crucial information. ChatGPT has demonstrated capability of GenAI (Generative Artificial Intelligence)-powered when deployed answer crisis-related questions in timely cost-efficient manner, thus replacing humans crisis communication. However, public reactions such messages remain unknown. To address this problem, study recruited participants (N1 = 399, N2 189, N3 121) conducted two online vignette experiments qualitative survey. The results suggest that, organizations fail handle requests, stakeholders exhibit higher satisfaction lower responsibility attribution instructing (vs. adjusting) information, as they perceived be more competent. satisfy that provide adjusting information) lead due competence. second experiment involving emergency scenario reveals regardless information provided (instructing or adjusting), greater positive attitudes toward high-competence low-competence) chatbots. further confirms experimental findings offers insights improve These contribute literature by extending situational communication theory nonhuman touchpoints deeper understanding using through lens machine heuristics. also practical guidance for strategically integrate human agents management based on context.

Language: Английский

Citations

8