Self-refining targeted readings recommender system
Muhammad Irfan Malik, M. Junaid Majeed, Muhammad Taimoor Khan, Shehzad Khalid · 2016
Huge volume of content is produced on multiple online sources every day. It is not possible for a user to go through these articles and read about topics of interest. Secondly professional articles, blog and forum have many topics discussed in a single discussion. Automatic knowledge-based topic models is a recent approach in Natural Language Processing that extract high quality topics from a large collection of documents. The quality of topics is improved through the model's auto-learning mechanism. In this paper, targeted reading content problem is addressed through automatic knowledge-based topic models, as a readings recommender system. The application recommends text documents based on contextual relevance. The learning module helps the model to learn certain rules from each recommendation, in order to recommend more relevant content in future. The contribution of this research work is to augment knowledge based models with contextual recommender systems. An application is developed that recommends targeted readings to the audience while the knowledge-based learning module grows in experience to serve the future users better.