In the era of digital economy, data is an important strategic asset for Internet enterprises, but the improvement of data asset valuation methods is also an urgent problem. This article establishes a valuation model that combines dual precision neural network algorithm and real option method. The model predicts future profits through dual precision neural network and constructs a data asset value evaluation model that combines user value model and B-S real option model based on the characteristics of data assets. Tencent Music is taken as an example for data asset value evaluation.