Semantic analysis based content recommendation system for e learning using feature selection clustering summarization and cnn

Abstract

Content recommendation system is most important for everyone due to newlinethe availability of huge volume of online resources. The content newlinerecommendations are applying in various fields including online healthcare, newlinee-commerce and e-learning. It is necessary today for leading the life newlinesuccessfully with the help of internet resources. The semantic analysis is newlinerequired to know the exact meaning of each term available in the material and newlineto be identified from large contents. The semantic relevant contents are to be newlinegrouped by applying data summarization and clustering processes. Apart from newlineall these, the data pre-processing is a basic need to perform well on clustering newlineand classification processes. For fulfilling the current requirements, this newlineresearch work is proposed a content recommendation system with the newlineincorporation of semantic analysis aware data-preprocessing, clustering, data newlinesummarization and classification to recommend the most suitable contents to newlinethe learners. First, new feature selection algorithms such as semantic newlinesimilarity incorporated feature selection and Fuzzy Decision Tree and newlineWeighted Gini-Index based Feature Selection Algorithm (FDTWGI-FSA) newlinehave been proposed in this work for selecting the relevant features. Second, newlinethis work applies an existing optimization algorithm called Fuzzy logic based newlineBee Swarm Optimization is used for optimizing the data. Third, the existing newlineclustering methods namely K-Means clustering and subtype fuzzy clustering newlineare applied to group the content. Fourth, a new semantic analysis aware data newlinesummarization technique is introduced to summarize the content. Fifth, the newlineexisting Enhanced Multiclass Support Vector Machine and Convolution newlineNeural Network (CNN) are used to classify the data or content. newline