Existing deep learning-based lithology identification methods encounter several critical bottlenecks: they fail to achieve robust model transfer in source-free scenarios, underutilize the intrinsic topological information of data, and are susceptible to noisy pseudo-labels. To address these challenges, a novel framework termed graph structure consistency and pseudo-label guided source-free domain adaptation (GCPG-SFDA) for lithology identification is proposed. This framework integrates graph-based modeling with the mean teacher framework to build a multi-dimensional system for feature optimization and knowledge transfer. Specifically, tabular lithological data is first transformed into graph structures and processed via graph neural network, enabling the extraction of high-order semantic relationships while preserving original features to enhance target-domain representation. Furthermore, three relationship graphs (teacher graph, student graph, and teacher-student graph) are designed. Graph consistency constraints optimize sample similarities in the feature space, improving the clarity of target domain classification boundaries. Meanwhile, a self-supervised exploration mechanism is implemented to encourage robust feature learning through structural perturbations and teacher-student output synchronization. Comprehensive evaluations indicate that GCPG-SFDA achieves superior performance in lithology identification, offering a robust solution for data-constrained geological tasks.

Understanding consumer preference for products or services is important for users (individuals, platforms, merchants, and so forth) to make decisions. However, the preference is difficult to observe. Based on the online ratings of consumers, we convert the ratings into pairwise comparisons and present an online optimization model to derive the ranking orders of the products or services. We employ a continuous Exp strategy to develop a learning algorithm to solve the online optimization problem, which has almost the same performance as the best strategyexpost. This approach cannot only handle dynamic rating information with arbitrary rating distribution but is also efficient in computation. We also investigate the impact of the learning rate on the ranking order and provide a real-world application of a recommendation system for illustration.