Feature engineering with gray wolf algorithm and fuzzy methods for Friend recommender system in social networks

Due to the expansion of the use of social networks, new areas of research in this field have been presented to researchers. One of these areas is using intelligent methods for friend recommender system. In this research, by using fuzzy methods and the gray wolf optimization algorithm, a solution for friend recommender system in social networks has been proposed. The use of fuzzy methods is considered to extend the extracted features in the network. The gray wolf algorithm has also been used to identify the appropriate subset of the feature set. Also, the process of learning the patterns in the extracted feature set has been done by the neural network method. The results of the implementation of this research and its comparison with other available methods showed that the artificial neural network was a good choice for choosing the learning model. The results showed that the feature selection mechanism using the gray wolf algorithm and the use of fuzzy information has a significant impact on improving system performance. In addition, the study of system performance on different data sets showed that the proposed method is highly accurate.