Decoding ChatGPT: A Taxonomy of Existing Research, Current Challenges, and Possible Future Directions DOI Creative Commons
Shahab Saquib Sohail, Faiza Farhat, Yassine Himeur

et al.

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Chat Generative Pre-trained Transformer (ChatGPT) has gained significant interest and attention since its launch in November 2022. It shown impressive performance various domains, including passing exams creative writing. However, challenges concerns related to biases trust persist. In this work, we present a comprehensive review of over 100 Scopus-indexed publications on ChatGPT, aiming provide taxonomy ChatGPT research explore applications. We critically analyze the existing literature, identifying common approaches employed studies. Additionally, investigate diverse application areas where found utility, such as healthcare, marketing financial services, software engineering, academic scientific writing, education, environmental science, natural language processing. Through examining these applications, gain valuable insights into potential addressing real-world challenges. also discuss crucial issues trustworthiness, emphasizing need for further development areas. Furthermore, identify future directions research, proposing solutions current speculating expected advancements. By fully leveraging capabilities can unlock across leading advancements conversational AI transformative impacts society.

Language: Английский

Privacy Aware Question-Answering System for Online Mental Health Risk Assessment DOI Creative Commons
Prateek Chhikara, Ujjwal Pasupulety,

John C. Marshall

et al.

Published: Jan. 1, 2023

Social media platforms have enabled individuals suffering from mental illnesses to share their lived experiences and find the online support necessary cope. However, many users fail receive genuine clinical support, thus exacerbating symptoms. Screening based on what they post can aid providers in administering targeted healthcare minimize false positives. Pre-trained Language Models (LMs) assess users' social data classify them terms of health risk. We propose a Question-Answering (QA) approach risk using Unified-QA model two large datasets. To protect user data, we extend by anonymizing training process differential privacy. Our results demonstrate effectiveness modeling assessment as QA task, specifically for use cases. Furthermore, model's performance decreases less than 1% with inclusion The proposed system's is indicative promising research direction that will lead development privacy-aware diagnostic systems.

Language: Английский

Citations

2
A comparative evaluation of machine learning and deep learning algorithms for question categorization of VQA datasets DOI
Deepak Suresh Asudani, Naresh Kumar Nagwani, Pradeep Singh

et al.

Multimedia Tools and Applications, Journal Year: 2023, Volume and Issue: 83(19), P. 57829 - 57859

Published: Dec. 13, 2023

Language: Английский

Citations

2
ZhongJing: A Locally Deployed Large Language Model for Traditional Chinese Medicine and Corresponding Evaluation Methodology: A Large Language Model for data fine-tuning in the field of Traditional Chinese Medicine, and a new evaluation method called TCMEval are proposed DOI
Jinyang Zhu, Qing-Yue Gong, Chunfang Zhou

et al.

Published: Oct. 20, 2023

The success of ChatGPT has showcased the potential applications Large Language Models (LLMs) in field Traditional Chinese Medicine (TCM), encompassing areas such as medical diagnosis, adjunctive therapy, and TCM talent cultivation. However, current challenges, including hardware constraints, insufficient model domain knowledge, difficulties domain-specific evaluation, have constrained fusion LLMs with TCM. In an attempt to address these issues, this paper introduces ZhongJing, a LLM fine-tuned within TCM, capable generating responses at rate 8 tokens per second, smoothly operating on local personal computers. To assess model's expertise, TCMEval evaluation method, designed concerning students' exams. Experimental results demonstrate that ZhongJing achieves 6.49 Score improvement over Chinese-LLaMA2 indicating ability generate more specialized compared baseline models.

Language: Английский

Citations

2
Conversational Artificial Intelligence in Digital Healthcare: A Bibliometric Analysis DOI

P. R. Visakh,

Pooja Meena,

V. S. Anoop

et al.

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 723 - 734

Published: Jan. 1, 2023

Language: Английский

Citations

1
Decoding ChatGPT: A Taxonomy of Existing Research, Current Challenges, and Possible Future Directions DOI Creative Commons
Shahab Saquib Sohail, Faiza Farhat, Yassine Himeur

et al.

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Chat Generative Pre-trained Transformer (ChatGPT) has gained significant interest and attention since its launch in November 2022. It shown impressive performance various domains, including passing exams creative writing. However, challenges concerns related to biases trust persist. In this work, we present a comprehensive review of over 100 Scopus-indexed publications on ChatGPT, aiming provide taxonomy ChatGPT research explore applications. We critically analyze the existing literature, identifying common approaches employed studies. Additionally, investigate diverse application areas where found utility, such as healthcare, marketing financial services, software engineering, academic scientific writing, education, environmental science, natural language processing. Through examining these applications, gain valuable insights into potential addressing real-world challenges. also discuss crucial issues trustworthiness, emphasizing need for further development areas. Furthermore, identify future directions research, proposing solutions current speculating expected advancements. By fully leveraging capabilities can unlock across leading advancements conversational AI transformative impacts society.

Language: Английский

Citations

1