This paper uses visual multimedia technology to establish a model for civil aviation transportation information services. A training subset is generated with a random forest, and the strength and relevance indicators of the random forest are examined. The generated decision tree is run on the corresponding out-of-bag data to evaluate its classification ability and obtain its weights. The leaf nodes of the decision tree are normalized to calculate the out-of-bag estimation of classification strength to realize the application of visual multimedia technology in the civil aviation transportation information service system. The system scores after application are all 8 and above, and the highest can reach 10. The quality of its service efficiency and level is improved, and passengers traveling by air are satisfied.