Deep Edu: A Deep Neural Collaborative Filtering for Educational Services Recommendation

Abstract

In the modern world, people face an explosion of information and difficulty to find the right choice of their interest. Nowadays, people show interest in online shopping to meet their demands increasingly. For researchers and students, finding and buying the desired books from online shops is very tedious work. Recently Recommender System is an excellent tool to deal with such problems, but the Recommender System is suffering from multiple problems such as data sparsity, coldstart, and inaccuracy. To address these problems, we propose Deep Edu a novel Deep Neural Collaborative Filtering for educational services recommendation. A Deep Edu architecture consists of three parts of a Deep Neural Network model (such as input layer, a multilayered perceptron, and an output layer). The Deep Edu provides the following contributions: first, the users' identifier and books identifier features are mapped into N-dimensional dense embedding vectors, second, the Multi-Layer-Perceptron (MLP) takes the N-dimensional and non-linear features. To increase the performance of Deep Edu in all metrics, we proposed the advance Loss function. Equipped with the following, Deep Edu not only capable of learning the Ndimensional and non-linear interactions between users' identifier and books identifier, but moreover, it also considerably mitigates the cold-start, data sparsity, and inaccuracy problem. Over significant experiments performed on real-world good books dataset, the results show that Deep Edu's recommendation performance obviously outperforms existing Educational services recommendation methods.

TD-DNN: A Time Decay-Based Deep Neural Network for Recommendation System

“…These Deep Learning techniques have recently succeeded in many complex tasks, i.e., computer vision, natural language processing, etc. Therefore, researchers have begun to apply Deep Learning algorithms to recommending music [36], books [11], videos [16], medical [37], interfaces [38], etc. The first Deep Learning-based recommendation model was Neural Collaborative Filtering (NCF) [39], which targets the poor representation of MF in a low-dimensional space and replaces it with neural network architecture.…”

“…In recent years, Deep Learning (DL) [9] has made significant progress and achieved notable success in many sectors such as healthcare [10], cybersecurity [10], natural language processing [11], audio recognition [11], computer vision [12], etc. The promising capabilities of DL have encouraged researchers to use deep architecture for recommendation tasks [13,14].…”

Intelligent personalised exercise recommendation: A weighted knowledge graph‐based approach

“…Collaborative filtering‐based techniques are generally used to recommend exercises to students [17,22,23,28,29,33,34]. In particular, with the help of collaborative filtering, Segal et al [29] derived a group of students with a similar study behaviour as that of a target student and recommended exercises to the target student, based on the consensus of the group of students regarding the difficulty level of the exercises or based on the ratings to the exercises given by students with similar learning style [17].…”

“…In Reference [28], an exercise score matrix of students was decomposed to predict the score of a student on a specific exercise, and exercise recommendation was conducted on the basis of the prediction. Besides recommending exercises to students, collaborative filtering is also applied for the recommendation of other educational resources to students, such as MOOC applications [22,23] and books [33]. Although collaborative filtering is quite successful in the recommendation domain, it only considers the common features among students and ignores the differences in learners' own knowledgeability.…”

Advances and Trends in Real Time Visual Crowd Analysis

“…These methods are dominant crowd modeling methods. Deep learning based methods (DLMs): As compared to TMLMs, recently introduced DLMs brought a large improvement in performance in various visual recognition tasks [ 89 , 90 , 91 , 92 , 93 ]. The TMLMs are based on handcrafted features, whereas, DLMs are more engineered.…”