With the development of big data technology, data mining methods provide a new way to combine the study of fine art and philosophy. In this study, K-means clustering algorithm and support vector machine model are used to extract features and classify the dataset containing 2,713 samples of typical philosophical fine arts works in China and 3,671 fine arts works integrating philosophical ideas from abroad. The study shows that the performance of K-means clustering after feature extraction is significantly better than other algorithms, with a clustering accuracy of 89.16% and a profile coefficient of 40.4. In terms of philosophical-emotional feature mining, the Pearson’s correlation coefficient of the support vector machine model in the utility prediction is 0.676, and the average absolute error is only 0.113, which is superior to the comparative models such as CNN and LSTM. The study classifies works of fine art into four major philosophical clusters: existentialism, metaphysics, social criticism and discernment, and perceptual experience, and predicts and analyzes the affective characteristics of each type of work. A survey of the number of participants in the International Society for the Philosophy of Works of Art 2020- 2024 shows that there is a growing trend towards the integration of fine art and philosophy. The study provides data support and theoretical framework for deepening the combination of fine art creation and philosophical thought, which is of great significance in promoting the multidimensional development of art creation.