Introduction
Users often need help selecting relevant information from the vast volumes of content that is accessible to them. In eLearning, recommendation engines assist learners throughout their educational journey by basing their recommendations on the content of the learning materials and user behavior.
Many eLearning platforms assign specific activities to students, but recommendation engines offer personalized suggestions, ensuring a tailored approach that maintains educational rigor. This blog provides a brief overview of the types of recommendation engines, and dives deeper into one particular model which is currently trending.
Types of Recommendation Engines
Recommendation engines are categorized as follows based on the logic powering them:
Collaborative Filtering: This method assumes that users with similar past preferences will continue to have identical preferences. The system generates recommendations by gathering data from user ratings and identifying users or items with comparable historical ratings.
Content-Based Filtering: This method generates recommendations by focusing on item descriptions and user profiles. It is effective when item information is available, even if user information is limited.
Hybrid Filtering: Hybrid recommendation engines combine collaborative and content filtering to minimize their limitations and provide better recommendations.
Knowledge Graphs based Recommendation Engines: There has been a recent surge in interest in integrating knowledge graphs (KGs) as supplementary information in recommendation engines. KGs provide a deeper understanding of item relationships and user preferences, improving the quality and interpretability of recommendations.
Understanding Knowledge Graphs
Knowledge graphs are interconnected frameworks incorporating contextual signals such as prior knowledge levels, learning styles, and current learning objectives to offer personalized eLearning recommendations. By leveraging advanced techniques like knowledge graphs, recommendation systems can significantly enhance the personalization and effectiveness of eLearning platforms, ultimately improving learning experiences and training outcomes.
Using Knowledge Graphs in mon’k to Deliver Personalized Learning
mon’k is a future-ready SaaS-based AI-driven platform that hosts a holistic knowledge suite: Adaptive Learning, eBooks, an Audio Video Player, a Reader as a Service (RaaS), and Journals. mon’k Adaptive Learning uses knowledge graphs to enhance personalized learning by organizing and interconnecting information intuitively and adaptively to individual learners’ needs.
Knowledge graphs in mon’k Adaptive Learning, model and track a learner’s knowledge and skills over time. They estimate a learner’s proficiency in specific concepts or skills based on their interactions with adaptive courses, such as exercises, assessments, and activities.
mon’k – Redefining the Learning Experience
The following steps outline how the recommendation engine powering mon’k Adaptive Learning works:
Modeling Knowledge States: The learning process begins by estimating the learner’s knowledge state in various concepts or skills. Knowledge graphs analyze a learner’s past interactions, preferences, and performance to recommend the most relevant and beneficial content. For example, based on a pre-test score, mon’k Adaptive Learning presents only the sections/modules of a course that are new to the learner, saving time and enhancing the experience.
Updating Knowledge States: The system analyzes the learner’s responses to different questions. As the learner engages with the training content, knowledge graphs map the relationships between different concepts and skills to create personalized learning pathways for each learner.
Adapting Content: The algorithm selects and presents appropriate content to the learner based on knowledge state updates.
Iterative Refinement: The system continuously refines its estimation of the learner’s knowledge state as new data becomes available. The knowledge graph evolves and improves, incorporating new information and insights from user data, ensuring the learning experience remains current and relevant.
Let’s look at a small example to understand this process better. The learner first completes a pre-test module before starting a course. Based on the pre-test score, mon’k Adaptive Learning presents only the course modules that are new to the learner. This action saves the learner time and provides a better experience. After completing the course, the learner takes a post-test assessment. If the learner does not secure the required score to complete the course, mon’k Adaptive Learning recommends the necessary modules for the learner to study to successfully complete the post-test assessment.
Conclusion
To summarize, knowledge graphs were employed to enhance m’onk Adaptive Learning and deliver a personalized educational experience. Our expert team reviewed multiple whitepapers, constructed a proof of concept, and tested it with subject matter experts (SMEs) and existing clients. Subsequently, they developed and integrated it into the mon’k platform — a future-ready, comprehensive knowledge suite.
If you would like to learn more about mon’k and its capabilities, write to us at marketing@impelsys.com.
Authored by: Abhishek Kumar
