<p>Multi-view clustering aims to integrate complementary and consistent multi-view information to uncover the data structure and assign samples into appropriate clusters. However, reducing data dimensionality while preserving view-specific features and extracting structural information within each view remains challenging. Facing this challenge, we propose a multi-view clustering method to enhance the clustering performance. The method constructs a multi-view non-negative matrix tri-factorization framework to reduce data dimensionality while preserving key features of the original data. Then, dual-graph constraints are employed to preserve the intra-view structural information. Next, pairwise co-regularization is applied to capture inter-view similarities and complementary information. Finally, these components are integrated into the clustering framework, with an automatic weighting strategy determining the weight of each view. An effective iterative update scheme is designed to solve the optimization problem of the proposed method, with the convergence of the update rules theoretically guaranteed. Extensive experiments on 12 real-world datasets demonstrate the effectiveness of the proposed algorithm, through both self-evaluation and comparison with 13 classical and state-of-the-art methods.</p>