Nowadays, when the existing algorithms are fused in the feature function network feature, they do not consider the different effects of different features on the saliency detection task. In order to solve this problem, this paper proposes a saliency feature fusion network based on multi-focus mechanism to complete the saliency detection task. While recognizing the important role of the network, it must also be realized that because most network users do not understand the details of network composition and network system analysis, the security of computer network data is relatively low. Unscrupulous hackers use their network technology and software to invade unauthorized networks and hosts without authorization. Therefore, network data security is very important. The competition and monopoly of the digital sharing economy platform has always been a hot topic of increasing attention from domestic and foreign economists. The platform economy is based on an efficient data collection and transmission system, a digital platform supported by continuously developing computing power and powerful data processing algorithms, which greatly promotes the development of social productivity. The core trend of the development of China's digital economy is the ecologicalization of the digital sharing economy. Small industrial platforms rely on massive user resources collected by a large platform to quickly connect the upper and lower channels of the industry to transform and form a new industry ecosystem.