Special issue on computational intelligence for social media data mining and knowledge discovery

The knowledge mining and discovery from social media data has been attracted by the industry and academia. It enables the enterprise and government to understand more knowledge hidden in these big data. Fortunately, the computational intelligence method can be developed to represent, analyze and extract useful information from these social media data. In this special issue, wehave 14 papers for this special issue. A summary of these papers is outlined below. In the paper titled “Supervised Shift k-Means Based Machine Learning Approach for Link Prediction Using Inherent Structural Properties of Large Online Social Network” by Praveen Kumar Bhanodia, Aditya Khamparia, and Babita Pandey, the authors propose a collaborative filtering-based approach to process complex and large social network data. Specifically, the approach exploits the node neighborhood features to extract similarity scores between nonlinked node pairs using the classical link prediction methods, then makes the final prediction of future links between node pairs by categorizing the feature network into clusters. The experimental results show that the proposed approach demonstrated an efficient grouping of social networks based on the potential of links between nodes of the network compared with classical techniques. In the paper titled “Distance Dynamics based Overlapping Semantic Community Detection for Node-Attributed Networks” by Heli Sun, Xiaolin Jia, Ruodan Huang, Pei Wang, Chenyu Wang, and Jianbin Huang, the authors propose a distance dynamics based overlapping semantic community detection algorithm (DOSC) for node-attribute networks. Specifically, the method is divided into three phases: First, they detect local single-attribute subcommunities in each attribute-induced graph based on the weighted vertex interaction model. Then, a hypergraph is constructed by using the subcommunities obtained in the previous step. Finally, the weighted vertex interaction model is used in the hypergraph to get global semantic communities. Experimental results in real-world networks demonstrate that DOSC is a more effective semantic community detection method compared with state-of-the-art methods. In the paper titled “Estimating Uncertainty in Deep Learning for Reporting Confidence to Clinicians in Medical Image Segmentation and Diseases Detection” by Biraja Ghoshal, Allan Tucker, Bal Sanghera, and Wai Lup Wong, the authors present an uncertainty estimation framework in deep learning for medical images by decomposing the uncertainty into two categories. Specifically, the framework to approximate Bayesian inference in deep learning by imposing a Bernoulli distribution on the incoming or outgoing weights of the model, including neurons. The experimental results show that the framework produces an equally good or better result in both quantified uncertainty estimation and quality of uncertainty estimates than approximate Bayesian neural networks in practice.