Animation Media Art Teaching Design Based on Big Data Fusion Technology

Animation, as an ancient art expression form, still has vigorous development, and the need for animation talents in society is increasing daily. This study first introduces the definition of animation and the development of animation at home and abroad. After that, the classification regression tree algorithm's principle and function theorem are described. This study divides the data into original and new animations based on big data fusion technology. It establishes a media art teaching system with search, recommendation, and playback as the three cores. Additionally, iteration is used to calculate the optimal hidden semantic matrix, a comparison is made between the benefits and drawbacks of the Sigmoid, Tanh, and ReLU functions, and lastly, the activation function chosen is the ReLU function. Compared with the loss value in the ideal case, the experimental findings comply with the likely criteria, and the categorical regression tree algorithm model predicts an error rate that falls within acceptable limits. Practically speaking, it is known that when the hidden factor dimension is 12, the system model works best for characterizing animation features. The comparison shows that the non-standard collaborative filtering recommendation system is inferior to the recommendations filtered by the categorical regression tree algorithm model. Following the use of the system, the students' drawing and directing abilities, animation scope, and animation appreciation level all improved significantly. The questionnaire survey concluded that the teachers and students of animation majors in universities were satisfied with the system.

@article{Wang2024,
  title     = {Animation Media Art Teaching Design Based on Big Data Fusion Technology},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {15},
  number    = {2},
  year      = {2024},
  publisher = {The Science and Information Organization},
  author    = {Rongjuan Wang and Yiran Tao},
  doi       = {10.14569/IJACSA.2024.0150275},
  url       = {https://doi.org/10.14569/IJACSA.2024.0150275}
}