Application of Semantic Analysis and LSTM-GRU in Developing a Personalized Course Recommendation System

Premalatha, M.; Viswanathan, Vadivel; Čepová, Lenka

doi:10.3390/app122110792

Cited by 6 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Artificial Intelligence technology plays a key role in blended e-learning platforms, particularly the use of deep learning, which analyses and processes large amounts of data to optimize the learning experience. This technology is able to accurately identify students' learning habits and preferences to provide more personalized recommendations in the online classroom [12][13]. Neural network models are an important component of deep learning.…”

Section: Neural Collaborative Filtering Algorithmmentioning

confidence: 99%

Application of Artificial Intelligence Technologies in Blended E-Learning Platforms

Liu,

Bai

2024

Advances in Intelligent Systems Research

View full text Add to dashboard Cite

With the development of AI technologies, AI application in education has become increasingly important, especially in blended e-learning platforms. However, these platforms often face the problems of course recommendation accuracy and data sparsity. To this end, the study proposes an improved neural collaborative filtering algorithm that incorporates temporal information to enhance the performance of the recommendation system. The results show that in explicit feedback evaluation, the research algorithm improves 58.3% in root-mean-square error and 59.1% in mean absolute error over the traditional collaborative filtering algorithm. In the implicit feedback evaluation, the proposed algorithm achieves 0.44 at NDCG@10 and 0.5257 at HR@10, both of which are superior to the traditional collaborative filtering algorithm. It can be found that the research algorithm can effectively improve the accuracy and adaptability of the recommender system. The study not only improves the course recommendation effect on the blended e-learning platform, but also provides an important reference value on how to optimize the recommendation system using temporal information and artificial intelligence technology.

show abstract

Section: Neural Collaborative Filtering Algorithmmentioning

confidence: 99%

Application of Artificial Intelligence Technologies in Blended E-Learning Platforms

Liu,

Bai

2024

Advances in Intelligent Systems Research

View full text Add to dashboard Cite

show abstract

“…Zhou et al ( 2022) introduced a time-aware recommendation system that considered learners' sequential enrolment data to recommend courses that matched temporal learner interests. Premalatha et al (2022) proposed a learner domain expertise model that analyzed the number of elective courses students completed and the grades they achieved.…”

Section: Learners' Previous Enrolment and Performancementioning

confidence: 99%

Extracting Course Features and Learner Profiling for Course Recommendation Systems: A Comprehensive Literature Review

Narimani,

Barberà

2024

IRRODL

View full text Add to dashboard Cite

As education has evolved towards online learning, the availability of learning materials has expanded and consequently, learners’ behavior in choosing resources has changed. The need to offer personalized learning experiences and content has never been greater. Research has explored methods to personalize learning paths and match learning materials with learners’ profiles. Course recommendation systems have emerged as a solution to help learners select courses that suit their interests and aptitude. A comprehensive review study was required to explore the implementation of course recommender systems, with the specifics of courses and learners as the main focal points. This study provided a framework to explain and categorize data sources for course feature extraction, and described the information sources used in previous research to model learner profiles for course recommendations. This review covered articles published between 2015 and 2022 in the repositories most relevant to education and computer science. It revealed increased attention paid to combining course features from different sources. The creation of multi-dimensional learner profiles using multiple learner characteristics and implementing machine-learning-based recommenders has recently gained momentum. As well, a lack of focus on learners’ micro-behaviors and learning actions to create precise models was noted in the literature. Conclusions about recent course recommendation systems development are also discussed.

show abstract

“…Moreover, AI-based CRSs extend their utility to suggest courses facilitating the acquisition of competencies relevant to students' future careers [37,38]. Conversely, some AI-based CRSs focus on optimizing academic outcomes by identifying courses with the highest probability of yielding favorable results and recommending them [39]. Additionally, AI-based CRSs explore students' latent interests by suggesting courses that are unexpected yet intriguing-a concept often termed serendipity [40].…”

Section: Ai-based Course-recommender Systemsmentioning

confidence: 99%

The Impact of AI-Based Course-Recommender System on Students’ Course-Selection Decision-Making Process

Cha,

Loeser,

Seo

2024

Applied Sciences

View full text Add to dashboard Cite

The course-recommender system (CRS), designed to aid students’ course-selection decision-making process by suggesting courses aligned with their interests and grades, plays a crucial role in fulfilling curricular requirements, enhancing career opportunities, and fostering intellectual growth. Recent advancements in artificial intelligence (AI) have empowered CRSs to deliver personalized recommendations by considering individual contexts. However, the impact of AI-based CRS on students’ course-selection decision-making process (inter alia, search and evaluation phases) is an open question. Understanding student perceptions and expectations of AI-based CRSs is key to optimizing their decision-making process in course selection. For this purpose, we employed speed dating with storyboards to gather insights from 24 students on five different types of AI-based CRS. The results revealed that students expected AI-based CRSs to play an assistive role in the search phase, helping them efficiently complete time-consuming search tasks in less time. Conversely, during the evaluation phase, students expected AI-based CRSs to play a leading role as a benchmark to address their uncertainty about course suitability, learning value, and serendipity. These findings underscore the adaptive nature of AI-based CRSs, which adjust according to the intricacies of students’ course-selection decision-making process, fostering fruitful collaboration between students and AI.

show abstract

Application of Semantic Analysis and LSTM-GRU in Developing a Personalized Course Recommendation System

Cited by 6 publications

References 29 publications

Application of Artificial Intelligence Technologies in Blended E-Learning Platforms

Application of Artificial Intelligence Technologies in Blended E-Learning Platforms

Extracting Course Features and Learner Profiling for Course Recommendation Systems: A Comprehensive Literature Review

The Impact of AI-Based Course-Recommender System on Students’ Course-Selection Decision-Making Process

Contact Info

Product

Resources

About