Detecting learning style and Knowledge level based on Felder Silverman model, Complexity theory, Bloom taxonomy, Ontology, and Literature based methodNorhan Kamal 1, Professor Mohamed Abo Rezka21School Of Information system, Arab Academy For Science Technology & Maritime Cairo, Egypt Abstract” In the past years building an adaptive e-learning environment was an attractive point to all research, and Nowadays this point has been tremendously developed to include not only adaptive e-learning environment, but also it will be depending on the intellectual dexterity and learning style of the learner as the course presentation will no longer be static but it will be established on the LS (learning style) of the learner.
Many researches have presented solutions for building an adaptive e-learning environment based on the student LS using Felder Silverman model either through questionnaire or by automatic detection through literature-based method, neural networks, etc. but the obstacle in using Questionnaire only was that the answers won’t be accurate as the student can just choose any answer to finish the questionnaire, and if it’s depending on the automatic detection only it will take time till the student interact with the provided system to be able to conclude the student behavior.
In our paper we will be combining both techniques with the help of the Ontology to detect the accurate student LS and detect the student knowledge through the complexity measurement of each question using six measures: A) external domain references, B) explicitness, C) linguistic complexity, D) conceptual complexity, E) level of difficulty F) intellectual complexity (Bloom level)  to detect the knowledge level of the student. Finally, our paper aims to eliminate the failure of students in any course as the student will either pass or drop the course and if the student passed, then he/she will have the required minimum knowledge from the course.KeywordLS, LSD, Learning style, Ontology, Literature-based method, FSLSM, Bloom Taxonomy, and Complexity.I. INTRODUCTION E-Learning is defined as interactive learning in which the learning content is available online and provides automatic feedback to the student’s learning. E-Learning systems are increasingly becoming a significant part of the strategy for delivering online and flexible learning . It’s important that in an Adaptive and personalized e-learning environment is to offer learning support than in traditional classroom along with offering personalized course presentation to the user . Preferences in learning differ from person to another. There are different learning styles that will suit each person some will prefer pictures, writing, verbal, etc… But for sure choosing the right learning style will help in confidence, motivating and accelerating the learning rate the questionnaire will be used as an assessment to retrieve information about the learning style, knowledge, and interest of the student. The aim was to produce a questionnaire for students, which could be addressed at the beginning . Later new automatic techniques were introduced to help in detecting the learning style more accurately than the questionnaire. Bayesian Network is one source of automatic detection ; (2) using Hidden Markov Models and Decision Trees . The most remarkable research on the second approach was conducted by Graf et al.  and finally Literature based method.II. MOTIVATION OF THE RESEARCH AND PROBLEM STATEMENTThe aim of this research is to build a dynamic e-learning environment personalized for each student based on our detection to the student learning style and knowledge level. Our research is significantly different than other previous researches for several reasons:Providing personalized e-learning profiles based on FSLSM using both Questionnaire and Literature based methods.An adaptive e-learning environment that’s personalized for each student using Ontology will provide a shareable view of the domain that will clear out the information structure that’s exchanged by the information system .Using complexity of the questions and Bloom Taxonomy in detecting knowledge level of the student.The Proposed model will help to detect the student learning style, knowledge level and guidance and suggestion of the best educational track that will suit the student style, skills, and knowledge.Most of the existent e-learning environment have fixed presentation for the course content regardless the student learning style, and if exists that the course is represented based on student learning style it will be either depending on the questionnaire or by monitoring the student behavior to suggest later on the presentation that suits the student style The proposed model won’t act on future presentation for course but will depend on presenting current course based on student learning style this will be done through initial detection using FSLSM Questionnaire combine with literature-based method and ontology to make sure that each part of the course was presented correctly to the learner, we will walk through the details in section IV. The rest of this paper is going to be organized as follows Section III Literature review and related works, section IV the proposed model in detail section V will be a conclusion and future work. III. LITERATURE REVIEW & RELATED WORKThis section will review the literature and related works for this study about the traditional LMS and its evolution with the evolution of web, e-learning and the widespread learning models and its detection methods, also will discuss the use of Ontology in ELearning. Finally, measuring the complexity of each question will be based on six measures: A) External domain references, B) explicitness, C) linguistic complexity, D) conceptual complexity, E) level of difficulty F) intellectual complexity (Bloom level) to detect the knowledge level of the student A. LEARNING STYLES Definition and concept of learning styles Learning styles are seen as characteristic cognitive, affective, and psychological behaviors that serve as relatively stable indicators of how learners perceive, interact with, and respond to the learning environment. LS points to an individual’s perfect way of handling and remembering new info and skills. According to Capretz (2006), each learning style has Pros and Cons, therefore, a person that will hold on to one style will never be a perfect learner (Moradkhan and Mirtaheri, 2011).B. FELDER-SILVERMAN MODEL:FSLSM is a learning style model that’s used frequently in improving learning in the traditional environment. [7 Examples for data sources that could be used for detecting the learning style are: log files, background data, and other personalization sources. C. LEARNING STYLES DETECTION:FSLSM articulates a different characteristic of learning, each dimension can be described as follows;  According to the way in which the learner assimilation the information, they might be demonstrated to sensing or intuitive learners, visual or verbal, active or reflective and finally, sequential or global depending on the way they recognize information.ILS (Felder-Silverman Indexing learning style) includes 44 questions, for each dimension there is 11 question. The questionnaire exist easily over the web  and offers scores as follow: 11A, 9A, 7A, 5A, 3A, 1A, 1B, 3B, 5B, 7B, 9B or 11B for every dimension of the four dimensions. The total score achieved by the student can be between 1-3, implication that the student preference is equally on the two dimensions of that scale if the score is between 5-7, this indicates that the student has a moderate preference for one dimension of that scale and will learn effortlessly in a teaching environment that empowers that dimension, if the score is between 9-11 implicates that the student has a strong preference for one dimension of that scale and the student will possibly has a big struggle in learning in an environment that doesn’t provide this preference. The symbols A, B indicates one pillar for every dimension in FLSM. Sabine Graf et al.  stated that ILS can be dependable as a method for detecting learning styles in LMS is a suitable instrument for assessing learning styles, and that the ILS is the best to permit individuals to compare the strengths of the qualified learning preferences rather than offering comparisons with other individuals.