Linked Data in E-Learning Platforms: Enhancement and Impact on Students Learning and Thinking

In the digital era, e-learning has emerged as a cornerstone of modern education, offering unprecedented flexibility and accessibility to learners worldwide. However, as the demand for online education has been grown, so challenges have been associated with effectively delivering quality learning experiences. One promising avenue for addressing these challenges lies in the integration of Linked Data technology within e-learning platforms.

The concept of Linked Data, has rooted in the principles of the Semantic Web, offers a powerful framework for organizing, connecting, and leveraging vast amounts of educational resources in a meaningful and interoperable manner. By structuring data in a way that facilitates semantic understanding and knowledge inference, Linked Data holds immense potential to revolutionize the landscape of e-learning. The current state of e-learning through the Semantic Web technology has been presented [1], with the objective of achieving two main goals: firstly, to discern the emerging themes in e-learning facilitated by Semantic Web utilization. Secondly, to underscore the obstacles and potentials presented by Semantic Web technologies in remote learning. Integrating the Citations’ Context and Reasons Ontology (CCRO) taxonomic hierarchy of reasons within a document has been proposed [2]. To streamline this process, they created a semantic annotation tool enabling authors to incorporate citation justifications while composing their papers. The package can be integrated into any authoring tool, and authors can tag a citation using appropriate CCRO properties. By doing so, a simple document can be transformed into a semantic document. The challenge of information security has been associated with the widespread implementation of online education during the COVID-19 pandemic, particularly within the framework of the Semantic Web [3]. Examination has delved into the necessity of steganography within this context, highlighting its significance despite the utilization of encryption methods. It has been revealed that they fall short of ensuring adequate safeguarding. Whereas Shahzad's research has delved into the ramifications of emerging e-learning technologies for university librarians and libraries. The study [4] has also addressed the obstacles that arise in the course of implementing these technologies.

Many applications of annotations are developed [5] in the literature for video annotation, such as multimedia annotation tools, video annotation systems, video annotation software, and video annotation learning systems [6]. The effectiveness of T-CoRe, a reflective tool, when used alongside annotated self-recorded videos and reflective writing, has been assessed [7]. The objective was to evaluate the application of digital technology in student teachers' teaching practices. The research results indicate a promising approach for teacher education, focusing on enhancing student teachers' proficiency with digital technology. The review [8] on digital video annotation tools, focusing on how video annotations were utilized in educational settings in various studies. It also highlights the professional competencies that were investigated and developed with the aid of video annotations, as well as the learning outcomes reported in the selected studies.

Video annotation is a technique for synchronizing commentary with video data. It is widely used in education and has received considerable research attention. However, they are diverse and dispersed. Ontologies for annotating educational resources can be utilized to improve information retrieval and search outcomes [9]. The study [10] on Social Annotation (SA) referred to a collaborative feature that allows users to highlight important text, make comments, and engage in discussions on the same online document. The purpose of the study was to explore how SA can be used to facilitate collaboration and knowledge sharing among users.

Research [11] aimed at automatically generating questions to facilitate learning about history using current linked open data (LOD) has been conducted. The purpose of this research is to elucidate LOD's potential as a learning tool. They established an open learning environment where students can access machine-understandable natural language content on a variety of subjects by connecting LOD to natural language texts.

Research [12] has concentrated on linked open data (LOD) in the educational area, creating RDF descriptions of video lectures taken from YouTube and Videolectures.net. An ontology derived from textbook data has been created to enhance the efficacy of learning materials [13]. In the realm of educational document retrieval, domain ontologies for the indexing of educational resources have been formulated. An IEEE LOM standard application profile has been suggested [14], with a focus on distant learning resources. It has been presented how this profile can be sufficiently general to be used with any kind of instructional resource and application. Additionally, an ontology model for this profile has been offered with the goal of enhancing the potential for educational resource retrieval and discovery in intelligent e-learning settings.

The study [14] offered a survey and examination of the literature on ontology-based recommenders for online education. Ontology-based e-learning recommenders employ many recommendation strategies, which they categorized. Along with classifying the categories of learning materials that e-learning recommenders suggest, they have also classified the knowledge representation method, ontology type, and ontology representation language employed by ontology-based recommender systems. The use of ontology for knowledge representation in e-learning recommender systems can enhance the quality of recommendations, as they demonstrated. They have also highlighted the future developments of this recommendation strategy in the context of e-learning. A subject or topic-based ontology was used to construct a portal for electronic learning. By focusing on the subject matter and searching the information semantically, one can retrieve content from the site [15]. The applications of semantic web technologies in e-learning have been categorized into five phases [16]: ontologies, instructional authoring ontologies, instructional assistance and adaptation, ontology-intelligent learning systems, and semantic web for social applications. Ontology and semantic annotation are included. The study [17] created an automatic test generation prototype that could produce multiple-choice questions based on domain ontologies. The first concern of the semantic web is the assertion of meaning. There are numerous free educational tools available on the web. Efficiently locating all relevant distributed educational resources is essential, students have difficulties maintaining their attention when watching videos and digesting the content.

Researches [18, 19] indicated that annotation tools can enhance student learning, with visual learning methods found to significantly boost student engagement. Education endeavors to cultivate scholarly engagement by prioritizing focused attention and concentration, particularly through video-based activities [20].

Various studies have demonstrated the effectiveness of video annotation as a tool for fostering reflection. This approach [21] prompts students to reflect on their academic endeavors. Students engaged in motivational and impactful self-reflection exercises that demanded deeper analysis [22]. Case study [23] highlighted the implementation of video-annotated projects aimed at enhancing advocacy communication skills among Juris Doctorate Students.

The crucial importance of video annotation for facilitating and motivating reflection among student instructors regarding their teaching methodologies was underscored [24, 25]. Additionally, peer feedback practices based on video are frequently used in classroom settings [26].

Another pivotal strategy involves course content recommendation, employing adaptive learning materials informed by ontologies and semantic guidelines. Numerous researches [27, 28] showcased ontology-driven approaches to personalize learning experiences according to individual learning styles. These studies have focused on tailoring course materials to suit learners' preferences and knowledge needs which is based on course ontology and user profiles [29].

Therefore, in this study, an e-learning platform was developed for El-Tarf University, along with a customized ontology. Video annotation was also integrated to enhance the learning experience. The platform provides online tools that allow students to semantically tag videos using linked data. The effectiveness of the online education process was evaluated through pre-tests, post-tests, and surveys administered to both students and teachers.

The following research questions are addressed to both teachers and students:

RQ1: How can multimedia content be annotated to improve the quality of video-based learning and ensure interoperability for both teachers and students?

RQ2: How can classes and teaching skills be enhanced by accommodating individual learning paces?

RQ3: How can students better comprehend concepts by using annotation tools while watching course videos?

RQ4: How can teachers effectively evaluate the outcomes of the proposed learning method?

The rest of the paper is organized as follows: Section 2 reviews existing literature relevant to our study. Section 3 details our contributions and the methodology employed in our research. Section 4 describes the platform that we developed and its associated semantic ontology. Section 5 evaluates the operation and impact of our e-learning system. Section 6 presents our findings and analysis. Finally, we conclude the paper with a summary of our results and their implications.

The aim of this study is to enhance e-learning by developing a platform for teachers and students, capable of generating a vast amount of question-and-answer content for online courses. Students engage with these courses, which leverage advanced multimedia technologies, enabling them to be evaluated and develop new skills and knowledge. The primary goal is to integrate semantic ontology technology into the newly created UCBET platform to improve learning processes and outcomes.

Our methodology involved analyzing foundational papers from relevant publications, defining research questions, and establishing research objectives. The first phase focused on platform development, incorporating custom ontologies and semantic annotation to enhance educational operations. A pre-test was conducted to assess students' prior knowledge, while a post-test evaluated the effectiveness of learning through semantically annotated videos. Additionally, an impact questionnaire was used to measure the influence of the platform on both teachers' and students' learning and thinking processes. The research methods and phases are detailed in Figure 1.

1.png

Figure 1. Research method and phases

All participants, in this study, provided informed consent before participating in the study. The consent process involved providing participants with detailed information about the study's objectives, procedures, and potential risks. Participation was entirely voluntary, and participants were informed of their right to withdraw from the study at any time without penalty.

To ensure the protection of participant privacy, all collected data were anonymized by assigning unique identification codes to each participant. No personally identifiable information was recorded or linked to the experimental data. Data were securely stored on password-protected systems with access limited to the research team.

By implementing these measures, we ensured that participant privacy and data security were rigorously protected throughout the study.

5.1 Group organization and data collection

The 80 participants were randomly assigned into two groups of 40 each (Group A and Group B) using a random number generator. This randomization process ensured that both groups were comparable in terms of demographic characteristics and baseline programming knowledge, thereby reducing the risk of selection bias. Participants were eager to learn the C programming language.  Students in group A were instructed and directed to study using our semantic video annotation, while students in group B were informed and guided to learn using video with web annotation.  The participants were selected based on their interest and willingness to learn the C programming language, ensuring motivation and engagement throughout the experiment. This selection criterion aligns with the study's objective of evaluating the effectiveness of the teaching methodologies for students genuinely invested in acquiring programming skills.

5.2 Use of the platform and pre-test

This phase consists of two key steps: First, students register for access to the platform. Second, they complete a pre-test designed to assess their existing knowledge. Administered before the course begins, the pre-test establishes a baseline understanding of the subject matter. In this context, the pre-test evaluates students' prior knowledge of the course content.

An experienced teacher developed the test based on the learning materials, specifically targeting students' understanding of the C programming language. The pre-test comprised ten multiple-choice questions, with a maximum score of ten. Participants received detailed instructions via email, ensuring they adhered to the learning schedule and respected the flexibility of the learning term. Table 2 outlines the pre-test questionnaire used to assess students' prior understanding of the C programming language.

Table 2. Pre-test questionnaire for students' prior understanding of the C programming language

Following the subscription, a statistical analysis was conducted to evaluate the differences between the two groups. The analysis used the covariance (ANCOVA) approach, with the pre-test scores serving as the covariate and the post-test scores as the dependent variables. This methodology was employed to establish the extent of disparity between the two groups and provide valuable insights to show the difference of the prior subject knowledge between group A and group B. The evidence has clearly indicated that both groups of students had an equal amount of prior knowledge. This is a crucial factor to keep in mind when evaluating the effectiveness of the activity as it ensures that any differences in performance between the groups can be attributed to the improvement itself and not to differences in pre-existing knowledge which are presented in Table 3.

Table 3. Descriptive data and ANCOVA of the pre-test results

5.3 Post-test to assess the student's intellectual progress

To truly put e-learning to the test. Students in group A were introduced to systems that focus on the connected data enrichment mechanism during the learning activity. Using commonly used vocabulary, this technique aims to connect annotations to external instructional datasets. They might look through the annotations in the documents using the semantic search paradigm. Group B has introduced learning with text annotation without using annotation-based linked data enrichment. When the learning activity has been completed, the post-test will be carried out on groups A and B. The post-test contains twenty questions expressed in pseudocode with four possible answers for each student who must select one answer. These post-test results for groups A and B are presented in Table 4.

The evaluation compares two types of learning: learning without semantic annotation and learning with semantic annotation, to establish the influence of linked data tools on learning and teacher motivation. The analysis revealed a significant difference in post-test scores between Group A and Group B (F = 3.65, p < 0.01). The experimental group, which used the semantic annotation approach, achieved significantly higher scores compared to the control group. This suggests that the semantic annotation approach was more effective in enhancing student learning outcomes than the approach without semantic annotation.

When discussing criteria such as “strongly agree,” “agree,” “disagree,” “strongly disagree,” and “undecided,” it is important to understand their role in gathering and interpreting opinions or feedback. These categories allow individuals to express varying levels of agreement or disagreement with a given statement, providing a nuanced view of collective sentiment. “Strongly agree” and “strongly disagree” capture the extremes of an opinion, indicating a high level of satisfaction. “Agree” and “Disagree” reflect more moderate positions, indicating a certain level of acceptance or disagreement without extreme intensity. “undecided” serves as a decisive option for those who feel neutral or lack sufficient information to form a final opinion. This spectrum helps analyze responses more accurately, allowing for more informed decisions to be made and it’s based on the feedback collected.

Table 4. ANCOVA results of the post-test of the two groups

In our study, we have observed that Group A had a very high concentration of individuals who strongly agreed. In contrast to, Group B’s opinions were more diverse, with responses split between strongly agree, agree, disagree, strongly disagree, and undecided. This indicates that Group A had a strong consensus on the proposed approach, while Group B showed a wider range of opinions and less homogeneity in their responses are presented in Figure 7.

7a.png

7b.png

Figure 7. Evaluation results for students in groups A and B

After incorporating video annotations that contained semantic content to enhance student learning engagement, both groups of students were asked to complete a post-questionnaire. Table 4 presents the ANCOVA result of the post-questionnaire ratings given by the two groups. The mean ratings for group A were 2.78 with a standard deviation of 1.09, while for group B, they were 2.24 with a standard deviation of 1.15. It was found that the post-questionnaire ratings of the two groups were significantly different, with a p-value of less than 0.01. The adjusted mean of group A (2.75) was significantly higher than that of group B (2.28). Therefore, it can be concluded that video annotation had a significant impact on facilitating collaboration and knowledge sharing among students.

5.4 Impact on the comprehension and thinking of students

In addition to the post-test, students completed an impact questionnaire to assess the level of cognitive gain. The questionnaire is presented in Table 5.

Table 5. Questionnaire for impact, comprehension, and thinking for students.

5.5 Impact for teachers

Questionnaire for teachers: The questionnaire was completed by nine professors. They had provided distance lessons using the university platform approach, and they were allowed to evaluate the method's success and impact on students' learning and thinking. The result of the questionnaire showed that all teachers strongly agreed to use this method. A questionnaire for teachers is presented in Table 6.

Table 6. Questionnaire for teachers.