LDA based integrated document recommendation model for e-learning systems

The increasing popularity of e-learning systems has created the necessity for the personalized recommendation model which can be used to optimize the effective learning environment for the learner. Personalized recommendation model is a specific type of information filtering system used to identify a set of objects that are relevant to a learner. Instead of a learner actively searching for information, recommender systems provide advice to learners about objects they might wish to examine. An integrated recommendation model would greatly help learners to find the most desirable documents in their fields of endeavor. Due to the textual nature of documents in corpus, content information could be integrated into existed recommendation methods. We propose a personalized integrated model for e-learning systems that consists of two steps: a) Using Latent Dirichlet Allocation topic modeling technique to make topic analysis on corpus, b) Introducing a similarity measurement to content-based recommendation approach.