Date Published: October 9, 2018
Publisher: Public Library of Science
Author(s): Sundus Ayyaz, Usman Qamar, Raheel Nawaz, Yong Deng.
A Recommender System (RS) is an intelligent system that assists users in finding the items of their interest (e.g. books, movies, music) by preventing them to go through huge piles of data available online. In an effort to overcome the data sparsity issue in recommender systems, this research incorporates a content based filtering technique with fuzzy inference system and a conformal prediction approach introducing a new framework called Hybrid Content based Fuzzy Conformal Recommender System (HCF-CRS). The proposed framework is implemented to be used in the domain of movies and it provides quality recommendations to users with a confidence level and an improved accuracy. In our proposed framework, first, a Content Based Filtering (CBF) technique is applied to create a user profile by considering the history of each user. CBF is useful in the situations like: lack of demographic information and the data sparsity problems. Second, a Fuzzy based technique is incorporated to find the similarities and differences between the user profile and the movies in the dataset using a set of fuzzy rules to get a predicted rating for each movie. Third, a Conformal prediction algorithm is implemented to calculate the non-conformity measure between the predicted ratings produced by fuzzy system and the actual ratings from the dataset. A p-value (confidence measure) is computed to give a level of confidence to each recommended item and a bound is set on the confidence level called a significance level ε, according to which the movies only above the specified significance level are recommended to user. By building a confidence centric hybrid conformal recommender system using the content based filtering approach with fuzzy logic and conformal prediction algorithm, the reliability and the accuracy of the system is considerably enhanced. The experiments are evaluated on MovieLens and Movie Tweetings datasets for recommending movies to the users and they are compared with other state-of-the-art recommender systems. Finally, the results confirm that the proposed algorithms perform better than the traditional ones.
Recommender Systems are playing an important role since 1990’s  by solving information overload problem and assisting users by making intelligent decisions in suggesting them items of their interest [2,3]. Recommendations can be of books , movies , music , clothes  etc. Today many online stores provide recommendations to users e.g., Amazon, Netflix, Youtube etc .
In literature, a variety of different techniques have been used for providing recommendations to users.
Hybrid Content based Fuzzy Conformal Recommender System (HCF-CRS) is focused on an effort to solve the sparsity problem and the lack of demographic data information, to predict ratings of the items and calculate the confidence level of each recommended item. The quality of recommendations provided to the user by HCF-CRS is improved by recommending only those items to the user which has confidence measure greater than the set threshold called the significance level ‘ε’.
In this section, the experimental evaluation of our proposed HCF-CRS is carried out.
Our proposed algorithms are tested on two datasets, MovieLens and Movie Tweetings which contains sparse ratings and it is experimentally proved that the conformal prediction framework works best with sufficient number of ratings while the content based fuzzy algorithm gives better results with the sparse dataset.
Recommender systems are widely adopted in several domains but is still contending with a challenge of data sparsity and unavailability of demographic data. The proposed HCF-CRS framework solves the data sparsity and lack of demographic information issue using the Content Based Filtering (CBF) technique, which creates a user profile by learning from the user history. A novel Fuzzy based technique is employed in our framework, which uses two features ‘similarity’ and ‘dissimilarity’ to compute the similarities and dissimilarities between the user profile and the movies data from the dataset to generate a list of users interested movies based on fuzzy rule set. HCF-CRS also initiates a conformal prediction technique to recommender systems by computing a non-conformity measure between the predicted ratings by proposed fuzzy system and the actual ratings from the movie dataset. The proposed Hybrid Content based Fuzzy Conformal Recommender System (HCF-CRS) generates a recommendation set with confidence level for each object to be recommended and the recommendation set is reduced with a confidence level higher than the set threshold. The reliability of the system is enhanced with the confidence being associated with each recommended item and thus the performance of the system is further improved. We have measured the performance of our approach using MAE, Precision, Recall and F-measure and compared with other state of the art recommender systems to validate its accuracy and reliability. For future work, the generation of highly personalized movie recommendations should be focused by taking into consideration other features such as context, trust or friendship. The accuracy or performance of the system can be improved by developing better non-conformity measurement method which adapt to recommender system. We can also consider, the speed (efficiency) of the system along with the accuracy.