Automatic Prediction of Learning Style Based On Prior Knowledge Using IRT and FSLM

With the rapid progress of online learning technology, e-learning environments offer an increasing number of learning resources and platforms. One standing problem of e-learning is delivering the same learning for all learners within a particular course without taking into consideration their individual learning needs. Ideally, the content must be adjusted to fit the individual learner's learning characteristics. To this end, this study proposes a framework for adaptive learning that can align course content with each learner's learning style. Given that the important rule of placement tests for assessing prior knowledge, we propose an automatic prediction learning style at the beginning of online learning session. To overcome the cold start issue, our placement test aims: (1) to principally evaluate students' pre-requisites and to (2) implicitly provide an extensive knowledge about their predominant learning style at the beginning of training. To do that, we used two methods: a placement test based on item response theory (IRT) and a rule-based method. To classify learners according to their learning style, we adopted The Felder-Silverman learning style model (FSLM) as the basis of classification for our proposed system.