Content area
Learning is a process of acquiring knowledge and skills in a particular domain of interest. This process, though continuous in one's life, ends with formative or summative assessment, in specific domain. In higher learning institutions, learning is acquired mostly through reading, experiments, observation, discussion, collaboration, lecturing etc, where sources of information play a fundamental roll. Repositories of learning resources are made available to students through portals that are integrated into e-learning management systems. Observations have showed that students are interested in these learning resources so that through utilising them they can acquire knowledge and skills, a part that is fundamental for one to be described as learned. The paper delves on the development and implementation of an ontology based innovative e-learning system with regard to acquisition of knowledge and skills by online learners at higher learning institutions set up where only core learner activities are identified. An e-learning ontology is used to model a domain of knowledge, thus complementing the functionalities of other learning systems that do not implement this technology. The paper presents assessment of students using assignments with instant feedback on the performance of the students following a Student-Teacher interaction algorithm (STia). The system is developed following creative design methodology and tested with the undergraduate students in the department of Computer Science at the National University of Science and Technology (NUST) in Zimbabwe. Empirical results are given with regard to the strengths and weaknesses of the system as seen by the users. Recommendations to higher learning institutions in terms of policy, e-learning technologies and utilisation of e-learning systems are given.
Abstract: Learning is a process of acquiring knowledge and skills in a particular domain of interest. This process, though continuous in one's life, ends with formative or summative assessment, in specific domain. In higher learning institutions, learning is acquired mostly through reading, experiments, observation, discussion, collaboration, lecturing etc, where sources of information play a fundamental roll. Repositories of learning resources are made available to students through portals that are integrated into e-learning management systems. Observations have showed that students are interested in these learning resources so that through utilising them they can acquire knowledge and skills, a part that is fundamental for one to be described as learned. The paper delves on the development and implementation of an ontology based innovative e-learning system with regard to acquisition of knowledge and skills by online learners at higher learning institutions set up where only core learner activities are identified. An e-learning ontology is used to model a domain of knowledge, thus complementing the functionalities of other learning systems that do not implement this technology. The paper presents assessment of students using assignments with instant feedback on the performance of the students following a Student-Teacher interaction algorithm (STia). The system is developed following creative design methodology and tested with the undergraduate students in the department of Computer Science at the National University of Science and Technology (NUST) in Zimbabwe. Empirical results are given with regard to the strengths and weaknesses of the system as seen by the users. Recommendations to higher learning institutions in terms of policy, e-learning technologies and utilisation of e-learning systems are given.
Keywords: Ontology, learning ontology, learning system, e-learning, online learners, learner activities
1. Introduction
From a pedagogical point of view, learning entails a learning space where learning activities occur whose main actors are lecturers/teachers/instructors and learners. Learning can be formal, informal, physical or virtual using available resources. Educators suggest that the learning environment should allow for flexibility, efficiency, intelligence, reusability and integration of content (Fouad, 2012). Teachers and learners engage in Internet related e-learning systems to acquire knowledge and skills from electronic-technologies and use them for educational development and problem solving. Current systems present multiple functionalities that are not necessary to learners and teachers hence there is need to scale down and concentrate on availing learning resources and students' assessments using ontology and semantic web technologies. This paper compliments the efforts rendered by related e-learning systems in the teaching and learning process and the proposed system could be an alternative for teachers and learners. Ontology and Semantic Web(SW) technologies are introduced so that search for content and problem solving become personalised following Student-Teacher interaction algorithm(STia).
2. Conceptualisation of e-learning systems
E-learning is the use of networked information and communications technology in teaching and learning entailing, online learning, virtual learning, distributed learning, networked and web-based learning with the following modalities; individualised self-paced e-learning online/offline and group - based e-leaning synchronous/asynchronous (Rubens, et al, 2011).
E-learning systems to solve specific problems have been developed and developers have envisioned the application of ontology and SW technologies in these systems. These technologies are still new in the field of research especially in teaching and learning. Learning standards such as Institute of Electronic and Electrical Engineers (IEEE), Learning Object Metadata Standard (LOMS), Instructional Management System (IMS), Sharable Content Object Reference Model (SCORM) and World Wide Web Consortium (W3C) are a requirement in e-learning systems (Ghaleb et al, 2009) and the later standard tries to adopt ontology and SW technologies to improve teaching and learning processes. The standard enables interoperability between learning content so that content developed using different tools can be handled by different e-learning systems and learning management systems. Currently, there is scarcity of domain ontology development for elearning systems, though it is noted that the field of medicine has made strides in developing ontologies (Bodenreider and Burgun, 2005). Numerous ontologies have been developed in biomedical community to represent biomedical terminologies in a common vocabulary for sharing and reuse across various fields but not in line with teaching and learning processes, hence the motivation in this research.
3.E-learning systems and Ontology
An ontology is defined as a formal explicit description of concepts in a particular or specific domain with their properties and equivalence relations (Aliman and Seghiouer, 2013) designating the building blocks for modelling the world, implying that there is theory that describes the objects in the world and that the world can be perceived by agents to perform some tasks. An e-learning ontology is used to model a domain of knowledge, thus complementing the functionalities of e-learning systems that do not implement this technology.
The application of ontologies and SW technology has a possibility in improving e-learning processes as depicted in the Language Technology for e-learning (TL4eL) project - Zurich University. This project implemented ontologies in the e-learning processes to enhance the management, distribution and retrieval of learning material within a Learning Management System (LMS) (Monachesi, et al, 2008). In this case, the concept of Intelligent Tutoring System (ITS) is presented as a system that uses Artificial Intelligence (AI) programming techniques and principles to imitate the expert model, student model and tutor model in the learning process where specific domain problems are solved but with no well-defined user needs, preferences, requirements and levels. Feedback to learners is also compromised. However, preliminary stage of adopting ontology and SW technologies is evident as there is ontological searches based on semantic definition, keywords and ontology browsing.
Ambikapathy (2011) proposed an intelligent e-learning system in accordance with current standards. The SW and the term ontology are defined implicitly including their applications. These concepts veil the introductory part of the SW and related technologies (Vanitha, et al, 2013). Also found in the same article is literature that discusses the influence of web technology on e-learning, including views of other researchers about e-learning 3.0, new developments in AI and its influence on the predicted e-learning 3.0. The concept of e-learning systems is understood as being sharing of knowledge and collaboration among learners mediated by teachers aided by Information Technology (IT) whose reasons for development are based on ubiquitous computing. This argument, follows the conceptualisation of e-learning systems in the field of AI (Rubens, 2011). Currently, elearning systems are characterised by challenges of information overload, non-semantic presentation of content and interoperability between systems or websites because most of them are not ontologically and semantically configured.
Mathematical relations and predicates can be employed in information presentation and retrieval processes, an idea that articulates well with the development of ontologies and the SW for e-learning systems (Ngwenya and Chilumani, 2013). The authors of this article, envisaged that e-learning systems were meant to improve on content quality and learning activities and that e-learning is almost becoming ubiquitous with the advancement of mobile technology. There is implementation of Learning Analytics (LA), a technique used to analyse learner's activities in a mobile learning set up. This is to enhance mobile learning or ubiquitous computing environment (Fulantelli, et al, 2013). Thus, e-learning systems focus on learning activities, cognitive structures and context of learning material with the learner at the centre stage (Fouad, 2012). Ontology technologies bridge the gap between keyword search and semantic-based search on the juvenile SW (Fernandez, et al, 2011). Ontologies have been developed for other domain but very little is related to educational processes. Researchers have called for the foundation in domain ontology design and development, knowledge assessment and computer -based learning (Chen, et al, 2011).
Some institutions have developed their proprietary systems with focus on the leaner and course management facilities, track online learning, manage classroom instructions, content delivery, regulatory compliance, competency, performance and course authoring. Student and teacher entities search for content to solve problems and acquire knowledge on particular domains. However, the systems that teachers and learners use present challenges, which needed an immediate attention for solutions in an e-learning environment. Evidently, these systems are not precisely and relevantly configured to meet user's needs, desires, requirements and preferences (Islam, et al, 2004). The same systems are used in assessment processes and students normally complain that these systems do not give instant feedback in terms of their performance and comments on answers. Learners argued that feedback constitute a paramount part in their learning and that they require access to repository of resources and assessment of their knowledge and skills in line with syllabi.
4.Methodological approach
E-learning system delves on software to compliment user needs and preferences hence the design based research becomes indispensable. Using this approach, focus is on the challenges faced by learners and teachers in search for educational content. Challenges faced by learners and teachers in the utilisation of current systems led to the understanding of the solution and subsequent development of the system, nesting scientific processes. In this case longitudinal approach was followed where questionnaires were used to gather data before the development of the system so as to understand user requirements. Then the system was developed following software principles and processes, finally testing the system and getting feedback from users. Later there was system visualisation using ontology, system design, development and subsiquently testing.
4.1System ontology visualisation graph
E-learning ontologies are modelled through two ways, thus it is either a frame based or Ontology Web Language (OWL) based. Figure 1 depicts an e-learning ontology visualisation graph for the proposed system modelled using OWL with classes, properties instances and reasoning.
Thing is the superclass, in reality this superclass could represent a university, a department or a faculty. Subclasses such as materials, instructor, students and course created based on the relationship between them and the superclass. This forms the foundation for the creation of relationships between objects, such as object properties that link objects and datatype properties that link objects to schema datatype or resource description framework rdf: literal. The other property created is the annotation property used to add annotation information to classes, individuals and other properties. The instances of materials are content and assessment. Content has instances such as tutorial, lecturer and notes, whereas assessment has tests, quiz, examination and assignments. For instructors, there is a teacher, tutor and lecturer but these can be used interchangeably and students can be pupils, postgraduate or under graduate students. Following this model is a reasoner over the system which is developed as one of the important components of the e-learning system.
4.2System Design
Systems development and implementation calls for principles and standard in e-learning environments. The system conforms to the following standards; Learning Object (LO), this standard enables access, interoperation and reuse of learning materials and content searches. The other standards incorporated in the system are IEEELOM(Learning object Metadata), Dublin Core, SCORM, IMS Question and Test Interoperability (IMSQTI), IMS Content packaging, learning Resource Metadata Initiative (LRMI). All these standards achieve interoperability, packaging of learning materials, usability and content search. The content in the e-learning system is represented following Fielding's Representation State Transfer (FEST) architectural style which prescribes a way for networked applications on the web to interact (Berners-Lee. and Cailliau, 2010). In this paper, the formative and summative assessment processes are ontologised and a novel process of issuing assignments and instant feedback is realised.
Ontology engineering involves the design and development of a domain or task ontology. It deals with the practicalities of implementing applied ontologies as software artefacts. Considering the fact that there are numerous ontologies that have been developed and are available online, it makes better sense to re-use these ontologies. The use of ontology design patterns and object oriented principles such as inheritance, aggregation, composition and association is important when considering re-use. Orłowski et al (2010) states that an ontology is a mechanism for controlling the precision of information as it goes through processing from input to output. Therefore, the task at hand is to identify relevant e-learning ontologies that are extendable especially to build knowledge in a knowledgebase useful for assessment tasks. An initial point was to extend OntoEdu's content ontology and activity ontology which is an Education Ontology in the OWL that involves activity and material ontology (Suteja et al, 2009).
When an assessment is created, the terms in the assessment populate a knowledgebase as its instances that commit to the teacher's ontology which is subset of the Education Ontology. When the student responds to the assessment, the terms found in their answer should correspond or match those of the assessment. The highest scores are attained when a student's inferred ontological structure and terms in the content of the response are a duplication of the teacher's ontology. Figure 2 shows the system architecture of the assessment system that permits instant feedback by using a STia that allows for student evaluation.
Students and lecturers engage in exchange of ideas in the process of learning and teaching. Teachers are the domain experts who impart knowledge and skills to the students. Teachers are the embodiment of the ideal domain ontology which the students are expected to download and reformat their own personal ontologies to replicate the ideal domain ontology. Learning assessment is conducted as a process of verification and validation of the extent to which the students' ontology matches the teacher's ontology. The grade obtained by the student after an assessment is a measure of alignment of the student's ontology to the teacher's ontology. The higher the score of the student mark the closer the similarity between the student's ontology and the teacher's ontology. The lower the score of the student mark the further away the resemblance between the student's ontology and the teacher's ontology.
The teacher creates an assignment in the form of multiple choice questions, the student takes the questions within a designated time frame (set by the teacher in minutes) after which the student can see the result of the test. The student for each test taken can view their score, the answers they submitted, the correct/wrong answers and the explanation for each. (explanation is written by the teacher).The system also permits the teacher to upload notes and learning resources in the form of portable document format (pdf), Microsoft word documents (doc, docx) and plain text documents (txt).
A web based user interface allows the student and teacher to access the system. This user interface is dynamically generated by a user agent according to the user's profile. The system follows a Model-ViewController (MVC) architecture. The model is an e-learning ontology which specifies the relationships and interactions among the entities in the system. The Views are generated dynamically by user agents from the ontology model. These Views are generated as Hyper Text Mark-up Language 5 (HTML5) files compatible with modern desktop and mobile phone web browsers. The Controllers are the user agents that map the Model to each user's View. The Controllers are written in PHP and Protocol and RDF Query Language (SPARQL). The system stores all entities of users, tests, questions, answers, submissions, explanations, and scores in an Apache Jena Fuseki Triple Store (RDF store). These records are in the form of Resource Description Framework (RDF) triples.
4.3STia algorithm
An ontologised STia is used to create an assessment in conjunction with an e-learning ontology that is used to encode knowledge in the domain of Computer Science. Students are assessed using assignments with instant feedback on their performance following STia. Imperial results on the strengths and weaknesses of the system indicate that students are interested in the systems that educate them based on their performance so that their efforts are directed at achieving quality education.
Assignments with instant feedback assist the students in realising the objectives of the learning processes better than without or delayed feedback. Rupere et al (2013) concur that feedback enhances education as shown in their results from an ontology based dictionary for language assessment of primary education. In this case, it is propound that when students receive feedback, they are able to determine their proximity in terms of the distance between their personal ontology and the ideal ontology as embodied by the teacher and communicated in the learning process. Feedback in assignments assists students by enabling them to measure their alignment or allowing them to reduce the noise that distorts the alignment of their own ontology to the ideal ontology. In determining the learning process and outcomes, there are three ontology alignment states for which feedback is intended to inform; firstly, identical conceptualisation elements, secondly dissimilar conceptualisation elements and thirdly, null conceptualisation elements.
In the first instance whereby there is high frequency of identical conceptualisation elements, the student obtains high grades or scores. This represents the scenario whereby the terms and their associated rules in the student's ontology and the teacher's ontology contain identical conceptualisation elements. During formative assessment, the interpretation of the feedback system implies that the vertical (that is to say the sub_class_of or is_a) branch of the student's ontology is updated with equivalent or horizontal (attribute_of or has_a) terms with their associated rules as in the teacher's ontology. Summative assessment implies a comparison of the whole structure of the student's ontology versus the structure of the teacher's ontology relevant to the knowledge and skill acquired in the learning resources.
In the second case of dissimilar conceptualisation elements, the student's ontology diverges from the teacher's ontology. In this case, the student feedback returns mid-range scores depending on the extent to which the divergence occurs. In this scenario, student's ontology has terms and associated rules that depart from those held in the teacher's ontology. The high end of the mid-range grades belong to divergence on the horizontal plane whereas the lower end represents divergence on the vertical plane. Unlike in the first case, the formative assessment refers to divergence in the subset of knowledge and skills currently being tested and summative assessment considers the whole set of knowledge and skill intended to be acquired from the learning resources. In the third case of null conceptualisation elements represents a failure in the learning process. This represents a scenario where the terms or rules in the student's ontology are not the same or do not belong to the set of terms and rules on the teacher's ontology. The codification of the algorithm is presented in table 1.
The uniqueness of this STia lies in the novelty of ontological philosophy of andragogy. The learning process is viewed in terms of personal or subjective ontology alignment to the ideal or objective ontology. The algorithm is applicable to both formative and summative assessment. The mark obtained by the student represents the grading of transformation extent of subjective ontology towards the objective ontology. The highest grades are assigned to high frequency of the occurrences of identical conceptualisation elements in the compared ontologies. The mid-range grades represent the measure of dissimilarity of ontology conceptualisation elements whereas the lowest grades belong to nullity of matching conceptualisation elements in the ontologies.
5.System evaluation
System testing was conducted with the part one undergraduate students at the National university of Science and Technology (NUST), with one of Department of Computer Science class of 20 students. Though the selection of students was purposive, it is in line with gender balance in education and Science, Technology, Engineering and Mathematics (STEM) requirements. These students were registered and they attended to the questions and interacted well with the system.
Learners highlighted that the STia system was the best because it presented work according to their requirements. They were able to get resources for reading and did their assignments where marking was done giving instant feedback to their performance. Those who rated STia best were 12 as compared to 3, 2 and 3 for Moodle, Sakai and Google classroom respectively. The institution introduced Moodle, followed by Sakai and later Google classroom but there was no clear policy with regard to adoption and utilisation. Learners observed many functionalities in these system, so when STia was introduced there was overexcitement on their part as they said some of these systems are not user friendly. However, they indicated that Google classroom can be better in terms of uptake. STia is scaled down, concentrating on availing learning resources and assessment using ontology and SW technologies.
Learners confirmed that STia is easy to use though they also appreciated Google classroom as indicated in Figure 4 above. Proportionally 9 students ascertained that notion and Google classroom seems to better than Sakai and Moodle as it is also new in the institution. The easiness of use was attributed to the fact that its interface was dynamically generated by user agent according to user needs, requirements and preferences.
Figure 5 shows frequencies in terms of what learners would prefer to use in the coming semester at the institution. STia could be one of the systems learners may use though Google classroom can be another option. Students indicated that there is no policy that govern the utilisation of these systems and apart from that no training was conducted especially for Moodle and Sakai. They said that training on Google classroom was conducted by lectures during lecture time hence its uptake. Though training for STia was not adequate, learners appreciated it. STia focusses on learning activities as expected by learners including context of learning materials with the learner at the centre stage.
Learners said that Google classroom and STia were better in terms of speed attributed to their memory and processing complexity and simplicity. They rarely crash as compared to Moodle and Sakai.
Students have been given the opportunity to evaluate the system. They are striking a balance in the utilisation of STia and Google classroom. These variations indicate that there is no clear policy on the utilisation of elearning systems at the institution. It is also noted that the systems are used for assignments and access to reading materials but without proper and established policy on how these can be utilised. None of them have been developed incorporating ontology and semantic technologies.
STia system was rated as good and very good and that more than half of the students are prepared to use it if granted permission in the subsequent semesters. This is so because students have pointed out that STia elearning system is good on assessment of the performance of students as it gives feedback immediately. This feedback is accompanied by answers to the questions, their scores as well as comprehensive explanation to the answers. Students are also able to access reading materials to consolidate their knowledge and skills. However, testing of STia was not very adequate considering the population and environment. A class of 20 students cannot be a good representation though results are favourable.
6.Discussion and conclusion
This paper has contributed ideas on attributes for consideration within e-learning systems to the theory of ontology interoperability and integration. Immerging from the discussion is the fact that the relationship between knowledge and content become explicit when using the ontology technology in searching, writing, and gathering, organising, developing of content and doing assignments. This would make teachers and students access content, collaborate, exchange ideas, and do homework and assignments in a personalised way. These technologies come with sharing and reuse of learning resources in different context and environments. Currently, there is not much that has been done in the area of domain ontology development for e-learning in specific subjects or courses. However, the SW and ontology technology have been exploited successful to some extent, though some promises by these technologies remain unaccomplished in other areas.
Recommendations are that content developers use available ontology technologies for teaching and learning to improve teaching/learning. Ontology and SW technology should be taken into account when developing elearning systems and that e-learning systems should be adopted and utilised with policy in place. This would cultivate and bring about the appreciation of ontology technology in e-learning systems. The other important future development would be to migrate e-learning sites to ontologically configured systems to facilitate teaching and content development processes.
References
Alimam, M. A. Seghiouer, H. (2013) Personalized E-Learning Environment based on Ontology to Improve Students' School Levels, International Journal of Computer Applications (0975 8887) Vol 84.12, [online], http://research.ijcaonline.org/volume84/number12/pxc3892977
Ambikapathy R. (2011) Ontology Based E-Learning System for a University on Semantic Web, International Journal of Mathematics and Computer Applications Research (IJMCAR), Vol 1.2, pp18-27. [online], http://www.tjprc.org/publishpapers/tjprcfile92
Berners-Lee T. and R. Cailliau, (2010) "World Wide Web: Proposal for a HyperText Project," tech. rep., CERN.
Bodenreider O. and A Burgun (2005), Biomedical Ontologies, Springer-Verlag;
Brinkley D., and R. V. Gotha, (2000), Resource description framework (RDF) schema specification 1.0. 2005. p. 211-236. W3C Candidate Recommendation, March 2000.
Chen B., Yu Lee C. and I. C. Tsai, (2011), Ontology - Based E-Learning System for Personalised Learning, 2011 International Conference on Education, Research and Innovation, IPEDR vol. 18 (2001), IACSIT Press, Singapore. www.ipedr.com/vol18/9-ICERI2011-R10008.pdf.
Fernandez M., Cantador I., Lopez V. , Vallet D. and P. Castells, (2011), Semantically enhanced Information retrieval: An ontology - based approach. Journal of Web Semantics, Volume 9, Issue 4, December 2011, pages 434 - 452
Fouad K. M., (2012), Proposed Approach to Build Semantic Learner Model in Adaptive E-Learning, International Journal of Computer Applications, Volume 58-Number 17, Foundation of Computer Science, New York, USA, BibTeX, November 2012.
Fulantelli G., Taibi D. and M. Arrigo, (2013), A Semantic Approach to Mobile Learning Analytics, First International Conference on Technological Ecosystem for Engineering Multiculturality, November, 14-15, 2013, Salamanca, Spain.
Ghaleb F., Daoud S., Hasna A., Alja'am J. M., El-Seoud S. M. and H. El-Sofany, (2006), E-Learning Model Based on the Semantic Web Technology. International Journal of Computing and Information Science, Vol. 4, No. 2 August 2006. www.ijcis.info/Vol4N2/pp63-71.pdf
Islam S. Md., Kumfuji S, Moura M and T. Hayana, (2004),Adopting Knowledge Management in an E-learning System: Insights and Views of KM and EL Research Scholars. An International Journal of Knowledge Management and Elearning, Vol 3, No 3, 375.
Monachesi P., Simov K., Mossel E., Osenova P. and L. Lemnitzer, (2008), What ontologies can do for e-learning, International Conference on Mobile and Computer aided Learning, 16-18 April 2008, Amman, Jordan, www.imclconference.org
Ngwenya S. and. Chilumani K. R, (2013), A Theoretical Semantic Web Framework for Developing Learning Content, ICEL/ICIT Conference Cape Town 2013, 8-12 December 2013.
Orłowski, C. Ziółkowski, A. and Czarnecki, A. (2010) Validation of an Agent and Ontology-Based Information Technology Assessment System, Cybernetics and Systems, 41:1, 62-74, DOI: 10.1080/01969720903408805, [online], http://dx.doi.org/10.1080/01969720903408805
Rubens N., Kaplan D. and Okamoto T., (2011), E-Learning 3.0: anyone, anywhere, anytime, and AI. In International Workshop on Social and Personal Computing for Web-Supported learning Communities (SPel 2011), Dec. 2011. www. activeintelligence.org/.../pdf/elearning-30-rubens-spel-2011-preprint.pd.
Rupere T, Chaka, P. Zanamwe, N. Mavhemwa P. M. (2013) An Evaluation of Language Ontology Web Based Assessment System, International Journal of Evaluation and Research in Education (IJERE), Vol. 2, No. 3, pp. 112~122, [online], http://www.iaescore.com/journals/index.php/IJERE/article/view/4445/pdf
Suteja, B. Istimewa, D. and Tengah D. J. (2009) Implementing Ontology for E-learning System Development In Indonesia based on Web 2.0, International Journal of Web Applications, Vol 1. 1, [online], http ://www. dline.info/ijwa/ful ltext/v1 n104
Vanitha K., Yasudha K, Dr. Sri Venkatech M. and Sowjanya K. N., (2013), Semantic E-Learn Services and Intelligent Systems using We Ontology. International Journal of Advanced Computer Science and applications, Vol. 4. No. 1, 2013. www.ijacsa.thesai.org.
Copyright Academic Conferences International Limited Jul 2018