Fatma Mohamed
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Fatma Mohamed has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Abdulhadi Shoufan
Department of Computer and Information Engineering, Center for Cyber-Physical Systems, Khalifa University, Abu Dhabi, United Arab Emirates
3 shared publications Abdulhadi Shoufan
Center for Cyber-Physical Systems (C2PS), Computer Science Department, Khalifa University, Abu Dhabi, UAE
2 shared publicationsBushra AlBelooshi
Department of Electrical and Computer Engineering, Khalifa University of Science and Technology, Abu Dhabi, UAE
1 shared publication- Chan Yeob Yeun
Department of Computer Science, Center for Cyber-Physical Systems, Khalifa University, Abu Dhabi, United Arab Emirates
1 shared publication - Ernesto Damiani
University of Milan, Italy
1 shared publication - Khaled Salah
Khalifa University, Abu Dhabi, United Arab Emirates
1 shared publication
Latest Publications
- Open Access
- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 4929-4939
MediCERN: An LLM-Powered Platform for Quality Evaluation of Medical Videos
YouTube has become a vital platform for disseminating medical information and guiding health decision-making. However, the literature confirms the prevalence of inaccurate, biased, or misleading content on this platform and underscores the need for robust quality assessment. This paper presents MediCERN, a platform that leverages a large language model (LLM) to evaluate the quality of medical videos using DISCERN, a widely accepted instrument for assessing health information reliability. Users can access MediCERN through a web interface to search for medical videos. The platform forwards each search query to YouTube and retrieves a list of relevant videos. MediCERN then checks whether any of these videos have been previously evaluated. Evaluated videos are presented to users ranked by their quality scores, while unevaluated videos are displayed without a score; however, users can request on-demand evaluation. The system processes these requests and provides scores within a few seconds. Sample analyses demonstrate substantial agreement between the model’s and experts’ ratings, with a Brennan–Prediger Kappa of 0.87. We further demonstrate the platform’s utility by analyzing over 5,000 videos. The results indicate that quality scores followed a Gaussian-like distribution, with most videos exhibiting medium quality. We also assessed the platform’s efficiency in terms of cost, time, and storage. For instance, the current prototype can evaluate one hour of video content in approximately 0.6 minutes at a cost of $0.0026 USD. These findings demonstrate that MediCERN offers a scalable and efficient solution for promoting access to high-quality medical content on YouTube.
- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 4929-4939
Open AccessMediCERN: An LLM-Powered Platform for Quality Evaluation of Medical Videos
YouTube has become a vital platform for disseminating medical information and guiding health decision-making. However, the literature confirms the prevalence of inaccurate, biased, or misleading content on this platform and underscores the need for robust quality assessment. This paper presents MediCERN, a platform that leverages a large language model (LLM) to evaluate the quality of medical videos using DISCERN, a widely accepted instrument for assessing health information reliability. Users can access MediCERN through a web interface to search for medical videos. The platform forwards each search query to YouTube and retrieves a list of relevant videos. MediCERN then checks whether any of these videos have been previously evaluated. Evaluated videos are presented to users ranked by their quality scores, while unevaluated videos are displayed without a score; however, users can request on-demand evaluation. The system processes these requests and provides scores within a few seconds. Sample analyses demonstrate substantial agreement between the model’s and experts’ ratings, with a Brennan–Prediger Kappa of 0.87. We further demonstrate the platform’s utility by analyzing over 5,000 videos. The results indicate that quality scores followed a Gaussian-like distribution, with most videos exhibiting medium quality. We also assessed the platform’s efficiency in terms of cost, time, and storage. For instance, the current prototype can evaluate one hour of video content in approximately 0.6 minutes at a cost of $0.0026 USD. These findings demonstrate that MediCERN offers a scalable and efficient solution for promoting access to high-quality medical content on YouTube.
- Open Access
- 22 March 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-12
Evaluating the quality of medical content on YouTube using large language models
YouTube has become a dominant source of medical information and health-related decision-making. Yet, many videos on this platform contain inaccurate or biased information. Although expert reviews could help mitigate this situation, the vast number of daily uploads makes this solution impractical. In this study, we explored the potential of Large Language Models (LLMs) to assess the quality of medical content on YouTube. We collected a set of videos previously evaluated by experts and prompted twenty models to rate their quality using the DISCERN instrument. We then analyzed the inter-rater agreement between the language models’ and experts’ ratings using Brennan–Prediger’s (BP) Kappa. We found that LLMs exhibited a wide range of inter-rater agreements with the experts (ranging from −1.10 to 0.82). All models tended to give higher scores than the human experts. The agreement on individual questions tended to be lower, with some questions showing significant disagreement between models and experts. Including scoring guidelines in the prompt has improved model performance. We conclude that some LLMs are capable of evaluating the quality of medical videos. If used as stand-alone expert systems or embedded into traditional recommender systems, these models can mitigate the quality issue of health-related online videos.
- 22 March 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-12
Open AccessEvaluating the quality of medical content on YouTube using large language models
YouTube has become a dominant source of medical information and health-related decision-making. Yet, many videos on this platform contain inaccurate or biased information. Although expert reviews could help mitigate this situation, the vast number of daily uploads makes this solution impractical. In this study, we explored the potential of Large Language Models (LLMs) to assess the quality of medical content on YouTube. We collected a set of videos previously evaluated by experts and prompted twenty models to rate their quality using the DISCERN instrument. We then analyzed the inter-rater agreement between the language models’ and experts’ ratings using Brennan–Prediger’s (BP) Kappa. We found that LLMs exhibited a wide range of inter-rater agreements with the experts (ranging from −1.10 to 0.82). All models tended to give higher scores than the human experts. The agreement on individual questions tended to be lower, with some questions showing significant disagreement between models and experts. Including scoring guidelines in the prompt has improved model performance. We conclude that some LLMs are capable of evaluating the quality of medical videos. If used as stand-alone expert systems or embedded into traditional recommender systems, these models can mitigate the quality issue of health-related online videos.
- Top Cited PapersOpen Access
- 3 January 2024
- Published by Springer Nature in BMC Public Health
- Vol. 24 (1) , 1-16
Users’ experience with health-related content on YouTube: an exploratory study
Background: This study focuses on health-related content (HRC) on YouTube and addresses the issue of misinformation on this platform. While previous research centered on content evaluations by experts, this study takes a user-centered approach and aims to explore users’ experiences with and perceptions of HRC videos and to establish links between these perceptions and some socio-demographic characteristics including age, gender, profession, and educational level. Methods: A quantitative research design was used in the study. 3,000 YouTube users responded to a 35-item anonymous questionnaire to collect information about the content they watch toward decision-making, their perceptions of the usefulness and bias of this content, what they identify as quality indicators for HRC, and what they recommend to improve the quality of such content on YouTube. The data were analyzed using descriptive statistics, frequency, and correlation analyses. Results: The results reveal that 87.6 percent (n=2630) of the participants watch HRC on YouTube, and 84.7 percent (n=2542) make decisions based on what they watch. Exercise and bodybuilding videos are the most popular, with over half of the participants watching them. 40 percent of the users watch YouTube videos to decide whether to consult a doctor or adopt specific health-related practices. In contrast to evaluations by experts in previous studies, most respondents perceive HRC videos on YouTube as useful and do not find connections between video quality and surface features like the number of views and likes. Weak or no correlations were observed between the perceived usefulness of HRC videos and age, gender, profession, or educational level. Participants’ recommendations for enhancing HRC quality align with previous research findings. Conclusions: Users turn to YouTube not only for health information but also as a decision-making tool. Combined with their generally positive attitudes towards content quality on this platform, this can have significant consequences for their health. Follow-up studies are needed to get more insights into decision-making behaviors and how users assess their decisions in retrospect.
- 3 January 2024
- Published by Springer Nature in BMC Public Health
- Vol. 24 (1) , 1-16
Top Cited PapersOpen AccessUsers’ experience with health-related content on YouTube: an exploratory study
Background: This study focuses on health-related content (HRC) on YouTube and addresses the issue of misinformation on this platform. While previous research centered on content evaluations by experts, this study takes a user-centered approach and aims to explore users’ experiences with and perceptions of HRC videos and to establish links between these perceptions and some socio-demographic characteristics including age, gender, profession, and educational level. Methods: A quantitative research design was used in the study. 3,000 YouTube users responded to a 35-item anonymous questionnaire to collect information about the content they watch toward decision-making, their perceptions of the usefulness and bias of this content, what they identify as quality indicators for HRC, and what they recommend to improve the quality of such content on YouTube. The data were analyzed using descriptive statistics, frequency, and correlation analyses. Results: The results reveal that 87.6 percent (n=2630) of the participants watch HRC on YouTube, and 84.7 percent (n=2542) make decisions based on what they watch. Exercise and bodybuilding videos are the most popular, with over half of the participants watching them. 40 percent of the users watch YouTube videos to decide whether to consult a doctor or adopt specific health-related practices. In contrast to evaluations by experts in previous studies, most respondents perceive HRC videos on YouTube as useful and do not find connections between video quality and surface features like the number of views and likes. Weak or no correlations were observed between the perceived usefulness of HRC videos and age, gender, profession, or educational level. Participants’ recommendations for enhancing HRC quality align with previous research findings. Conclusions: Users turn to YouTube not only for health information but also as a decision-making tool. Combined with their generally positive attitudes towards content quality on this platform, this can have significant consequences for their health. Follow-up studies are needed to get more insights into decision-making behaviors and how users assess their decisions in retrospect.
- Top Cited PapersOpen Access
- 28 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 125576-125599
YouTube and Education: A Scoping Review
YouTube has evolved to a global platform for formal and informal education. In contrast to traditional sources of learning multimedia, YouTube is a social platform with numerous characteristics that make its real value for education not obvious. We neither know how reliable the learning content on YouTube is, what best-practice strategies for using this platform in education are nor how watching YouTube affects students’ performance and behavior. To shed light on these questions, we conducted a scoping review of the literature on YouTube and education. A total of 647 publications were included and analyzed thematically. Four research themes could be identified: (1) Content creation and assessment (2) User attitudes and acceptance (3) Usage strategies and behaviors (4) Impact on student learning. The findings of the respective studies were analyzed and compiled theme by theme. The main results of this review are: (1) There is an increasing concern about content quality on YouTube. (2) Despite versatile production and usage strategies, no relationships were established between such strategies and learning. (3) Most studies on the impact of YouTube on student learning reported positive results in the form of enhanced skills, competencies, interest, motivation, engagement levels, or test performance. We conclude that YouTube is a rich, free, easy-to-use, and enjoyable source of learning content. However, the challenges and risks associated with this platform suggest that it is best suitable for guided learning where teachers make or select the content and include it in a well-defined, pedagogy-driven learning context.
- 28 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 125576-125599
Top Cited PapersOpen AccessYouTube and Education: A Scoping Review
YouTube has evolved to a global platform for formal and informal education. In contrast to traditional sources of learning multimedia, YouTube is a social platform with numerous characteristics that make its real value for education not obvious. We neither know how reliable the learning content on YouTube is, what best-practice strategies for using this platform in education are nor how watching YouTube affects students’ performance and behavior. To shed light on these questions, we conducted a scoping review of the literature on YouTube and education. A total of 647 publications were included and analyzed thematically. Four research themes could be identified: (1) Content creation and assessment (2) User attitudes and acceptance (3) Usage strategies and behaviors (4) Impact on student learning. The findings of the respective studies were analyzed and compiled theme by theme. The main results of this review are: (1) There is an increasing concern about content quality on YouTube. (2) Despite versatile production and usage strategies, no relationships were established between such strategies and learning. (3) Most studies on the impact of YouTube on student learning reported positive results in the form of enhanced skills, competencies, interest, motivation, engagement levels, or test performance. We conclude that YouTube is a rich, free, easy-to-use, and enjoyable source of learning content. However, the challenges and risks associated with this platform suggest that it is best suitable for guided learning where teachers make or select the content and include it in a well-defined, pedagogy-driven learning context.
- Open Access
- 11 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 51155-51166
Choosing YouTube Videos for Self-Directed Learning
YouTube provides a vital source for self-directed learning. YouTube’s search engine, however, ranks video clips according to popularity, relevancy, and view history rather than quality. The effect of this ranking on learners’ behavior and experience is not clear: Do learners tend to choose from the top of the returned search list? Does the choosing behavior affect their learning? Is the type of sought knowledge relevant in this process? To answer these questions, we conducted two experiments with sophomore-level students in electrical and computer engineering programs. The students were asked to learn about two new topics by watching YouTube video clips of their choice. The first topic conveys procedural knowledge about using the Quine McCluskey algorithm for minimizing logical functions. The second topic relates to the concept of the set-reset latch. In each learning session, the students had to report their watching behavior and experience by responding to an online questionnaire as well as to solve a problem related to the respective topic. The results show a clear tendency to choose from the top of the returned list. However, students’ performance in problem-solving was found to be uncorrelated with the choosing behavior. These results were similar for procedural and conceptual learning although the students’ performance in solving the conceptual problem was significantly lower. These findings indicate that university students who seek YouTube for self-directed learning can freely choose from the top of the returned search list without concern. There is no evident harm in doing so. However, students need to be thoughtful when using YouTube for conceptual learning. They should use different strategies such as watching multiple video clips, selecting video clips with higher viewer ratings, or watching video clips with related procedural knowledge to support the learning of new concepts.
- 11 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 51155-51166
Open AccessChoosing YouTube Videos for Self-Directed Learning
YouTube provides a vital source for self-directed learning. YouTube’s search engine, however, ranks video clips according to popularity, relevancy, and view history rather than quality. The effect of this ranking on learners’ behavior and experience is not clear: Do learners tend to choose from the top of the returned search list? Does the choosing behavior affect their learning? Is the type of sought knowledge relevant in this process? To answer these questions, we conducted two experiments with sophomore-level students in electrical and computer engineering programs. The students were asked to learn about two new topics by watching YouTube video clips of their choice. The first topic conveys procedural knowledge about using the Quine McCluskey algorithm for minimizing logical functions. The second topic relates to the concept of the set-reset latch. In each learning session, the students had to report their watching behavior and experience by responding to an online questionnaire as well as to solve a problem related to the respective topic. The results show a clear tendency to choose from the top of the returned list. However, students’ performance in problem-solving was found to be uncorrelated with the choosing behavior. These results were similar for procedural and conceptual learning although the students’ performance in solving the conceptual problem was significantly lower. These findings indicate that university students who seek YouTube for self-directed learning can freely choose from the top of the returned search list without concern. There is no evident harm in doing so. However, students need to be thoughtful when using YouTube for conceptual learning. They should use different strategies such as watching multiple video clips, selecting video clips with higher viewer ratings, or watching video clips with related procedural knowledge to support the learning of new concepts.
- 1 September 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 301-306
A Scattering Technique for Protecting Cryptographic Keys in the Cloud
Cloud computing has become a widely used computing paradigm providing on-demand computing and storage capabilities based on pay-as-you-go model. Recently, many organizations, especially in the field of big data, have been adopting the cloud model to perform data analytics through leasing powerful Virtual Machines (VMs). VMs can be attractive targets to attackers as well as untrusted cloud providers who aim to get unauthorized access to the business critical-data. The obvious security solution is to perform data analytics on encrypted data through the use of cryptographic keys as that of the Advanced Encryption Standard (AES). However, it is very easy to obtain AES cryptographic keys from the VM's Random Access Memory (RAM). In this paper, we present a novel key-scattering (KS) approach to protect the cryptographic keys while encrypting/decrypting data. Our solution is highly portable and interoperable. Thus, it could be integrated within today's existing cloud architecture without the need for further modifications. The feasibility of the approach has been proven by implementing a functioning prototype. The evaluation results show that our approach is substantially more resilient to brute force attacks and key extraction tools than the standard AES algorithm, with acceptable execution time.
- 1 September 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 301-306
A Scattering Technique for Protecting Cryptographic Keys in the Cloud
Cloud computing has become a widely used computing paradigm providing on-demand computing and storage capabilities based on pay-as-you-go model. Recently, many organizations, especially in the field of big data, have been adopting the cloud model to perform data analytics through leasing powerful Virtual Machines (VMs). VMs can be attractive targets to attackers as well as untrusted cloud providers who aim to get unauthorized access to the business critical-data. The obvious security solution is to perform data analytics on encrypted data through the use of cryptographic keys as that of the Advanced Encryption Standard (AES). However, it is very easy to obtain AES cryptographic keys from the VM's Random Access Memory (RAM). In this paper, we present a novel key-scattering (KS) approach to protect the cryptographic keys while encrypting/decrypting data. Our solution is highly portable and interoperable. Thus, it could be integrated within today's existing cloud architecture without the need for further modifications. The feasibility of the approach has been proven by implementing a functioning prototype. The evaluation results show that our approach is substantially more resilient to brute force attacks and key extraction tools than the standard AES algorithm, with acceptable execution time.