Development of a Content Based Recommender Using Dynamic Artificial Neural Network

Abstract: In recent years, we�ve been facing a boost in advancement of Artificial Intelligence (AI). It�s continuously setting benchmark of our thought level of which AI is possible. Although using Artificial Neural Network (ANN) to achieve AI is somewhat less explored area primarily because of complexity of the training process, AI is moving towards ANN approach recently with many sophisticated learning algorithms. In this paper, we tried to develop a content based recommender which can be used as a personal assistant for categorizing and making recommendation on news articles. Our primary goal was to evaluate the possibility of such a recommender which can be dealt with a multiple layer network which has dynamic number of input nodes. We were also evaluating the performance of the system under various configurations. We are calling the system dynamic artificial neural network (DANN).

Keywords: recommender; artificial intelligence; artificial neural network; content based recommender; dynamic artificial neural network.