Beyond Learning: Personalized Learning Environment
Personalized learning is an educational approach that aims to customize learning for each student's strengths, needs, skills and interests. Each student gets a learning plan that's based on what he/she knows and how he/she learns best.
By: Shatha Jararaa
Currently, this approach is a trend in the educational field and interests young professors around the world despite the fact that investment in this field is very limited. One of these Professors is Mona Demaidi, a Palestinian woman with great vision and big ambition.
Dr. Demaidi did her MSc in Software Engineering and Data Management at the University of Manchester, UK and PhD between Birmingham City University and the University of Manchester, UK with a specialization in Advanced Software Engineering. She joined An-Najah National University in 2016, to become the youngest female with a PhD certificate at the Faculty of Engineering and Information Technology. Her research focuses on building personalized learning environments using machine learning. In 2018, she published a book and two journal papers which introduces a new platform for auto-generating assessment questions using semantic web.
Personalized Learning Environments System
Dr. Demaidi produced two research papers titled: "OntoPeFeGe: Ontology-Based Personalized Feedback Generator" and "Evaluating the quality of the ontology-based auto-generated questions" as well as a book titled: "Terminological Ontology Evaluator in eLearning" from which she developed the idea of Personalized Learning Environments System. Dr. Demaidi believes that content management systems such as Moodle are regardless students' characteristics. In other words, Moodle offers all the material online where all students access 24/7. Students doing well and those who are struggling the course get the content in the same manner; therefore, popped the idea of creating a complete system from A to Z that focuses more on students as individuals.
Platform operation mechanism
The system provides different feedbacks to each student and has an auto-generating content that meets students' characteristics. It has more than one actor or as Dr. Demaidi calls them "users"; namely, teacher actor, student actor and the system. The system keeps tracking students' thought and interests such as what kinds of things he/she is actually reading. Then the teacher starts adapting feedback assessment questions. The auto-generation is not random but based on templates like Blooms taxonomy.
Knowledge graphs and ontologies
When we interviewed Dr. Demaidi she mentioned that the system is built on ontologies similar to the human brain. It takes the education material and builds over it knowledge graphs. Ontology is detailed knowledge graph that the computer could read and makes reasoning for it just like the voice command system. There are candidate ontologies where the developer can choose one based on students needs. Dr. Demaidi's research is still on going, but the staff finished two important phases; the experimental and the knowledge graph assessment tool. It is worth mentioning that Dr. Demaidi and Dr. Mohammad Jarrar from Birzeit University are the only professors in Palestine working on ontology.
The Role of the teacher in the system
Regarding the teacher role in the system, Dr. Demaidi says: "We can't take the teacher out of the educational process as he/she is essential to be part of the system. The teacher evaluates the feedback provided to the student and makes sure it suits him/her. Therefore, the teacher is a facilitator and evaluator not a controller."
The system doesn't only provide personalized feedback to the students, but also assists the teacher who takes a long time building the content and assessment questions. All the teacher has to do is to make sure that the content and the questions are addressing all students.
Dr. Demaidi talked about the challenges she faced while implementing the system, most important of which is the natural language processing. For example, if the system auto-generates a question that assesses the educational concept, the question format might not be clear to the students, so we still need an input from the teacher to make sure that the format is appropriate to each student. Another challenge is that educational ontologies worldwide are very limited and the existing knowledge graphs don't actually focus on the educational domain. That is, we don't know if the knowledge graph suits the students or not. She added that in Palestine and worldwide there is very limited investment in the educational field.
When we asked Dr. Demaidi about the application of the system at An-Najah, she answered: "The system needs further research for implementation, but I did an experiment for computer networks students at An-Najah. I chose this course as there is an existing knowledge graph for computer networks, but our staff enhanced it to suit our students." She added: "I published a research paper and wrote a book, but our computer engineering students did all the work and built the system."
Dr. Demaidi mentioned that teachers need training before utilizing the system. She suggested integrating the system with Moodle as a first step to create plugins with the E-Learning Center with Moodle supporting the plugin.
How the content is presented
The content presentation is still under progress as there are different types of learners; namely, visual, auditory, reading/writing and kinesthetic.
Moodle usage challenges
Dr. Demaidi said that she already uses Moodle in her courses, but it has no real time notification. That is, the students don't know that they have an assignment until they're told. In Moodle, teachers have to create everything from scratch and the feedback is not personalized. It is also not interactive, so teachers prefer to use Facebook as it has real time notification. Moodle is used all over the world taking a conventional learning style as it is a content management system in which all students regardless their differences receive the same content. It provides them with 24/7 access to the material or that's how teachers perceive it. She added: "We can add extra things to Moodle, but still it is not personalized."
Exams' correction is done automatically, but there is something called question quality. Question quality takes into consideration two measurements, difficulty level measurement and discrimination index. Difficulty level measurement measures how difficult the questions are for the students while discrimination index measures how the question can discriminate between good students and those struggling the course. Discrimination index also has two theories, the classical test theory which depends on the sample of the students and the item response theory. The item response theory is based more on machine learning technique not the sample of the students. Therefore, teachers can predict their students' needs.
Students' response to the experiment
Dr. Demaidi compared the students' response before and after the experiment. The results were shocking that students' response was highly positive. On the contrary, they had a negative impact from the conventional feedback. They said that the feedback was very good even though they didn't know that they were being evaluated individually. They also had a significant improvement in performance.
The system is based on the internet as it is essential in education. Students can stay at home, see the content they want and interact with it as they like.
Dr. Demaidi concluded: "In Palestine we have the talents and the minds, but we need guidance. If we invest in the minds, we'll reach out to the whole world. We should support each other and look at the bright side. Each of us has a message; mine is to work hard and have vision to make a change to the world which I consider part of our struggle. We should change the stereotyped image of Palestine because Palestine worth it."
Besides her academic work, Dr. Demaidi is an active member of several societies which globally support women in the technology sector. She has been working with Women in Engineering and Arab Women in Computing since 2014, and in 2017 became the first female supervisor of the Institute of Electrical and Electronics Engineers (IEEE) at An-Najah student branch.
Being a key part in such societies, she organized several initiatives which aim to empower women and reduce the unemployment rate in Palestine. In the technology sector in Palestine, the number of undergraduate females is high, however, in the tech industry, the unemployment rate for women is 60% which is three times more compared with males. Dr. Demaidi studied the reasons behind the high percentage and started applying several activities at the university and in the Palestinian community, to encourage women and help them improve their technical and communication skills. Currently, the IEEE committee at the university has more than 50%, female members. In addition, in most of the technical activities held by IEEE, females form more than 57%. Besides IEEE, Dr. Demaidi is working on encouraging women to attend regional conferences in the MENA region by helping them apply for scholarships to attend the Arab Women in Computing conference, the biggest conference in the region that celebrates and support women. She believes that exposing Palestinian undergraduate women to regional and international experiences is essential, as it helps them enhance their communication and networking skills. Moreover, attending such events will have a great impact on Palestinian society, as women will have equal job opportunities as males.
For further information, please visit the links: