Multi-view clustering (MVC) holds a significant role in domains like machine learning, data mining, and pattern recognition. Despite the development of numerous new MVC approaches employing various techniques, there remains a gap in comprehensive studies evaluating the characteristics and performance of these approaches. This gap hinders the in-depth understanding and rational utilization of the recently developed MVC techniques. This study formalizes the basic concepts of MVC and analyzes their techniques. It then introduces a novel taxonomy for MVC approaches and presents the working mechanisms and characteristics of representative MVC approaches developed in recent years. Moreover, it summarizes representative datasets and performance metrics commonly employed for evaluating MVC approaches. Furthermore, we have meticulously chosen 35 representative MVC approaches to conduct an empirical evaluation across seven real-world benchmark datasets, offering valuable insights into the realm of MVC approaches.

Multi-view clustering (MVC), which aims to explore the underlying structure of data by leveraging heterogeneous information of different views, has brought along a growth of attention. Multi-view clustering algorithms based on different theories have been proposed and extended in various applications. However, most existing MVC algorithms are shallow models, which learn structure information of multi-view data by mapping multi-view data to low-dimensional representation space directly, ignoring the nonlinear structure information hidden in each view, and thus, the performance of multi-view clustering is weakened to a certain extent. In this paper, we propose a deep multi-view clustering algorithm based on multiple auto-encoder, termed MVC-MAE, to cluster multi-view data. MVC-MAE adopts auto-encoder to capture the nonlinear structure information of each view in a layer-wise manner and incorporate the local invariance within each view and consistent as well as complementary information between any two views together. Besides, we integrate the representation learning and clustering into a unified framework, such that two tasks can be jointly optimized. Extensive experiments on six real-world datasets demonstrate the promising performance of our algorithm compared with 15 baseline algorithms in terms of two evaluation metrics.