An efficient way to handle the excessive number of user requests by the existing infrastructure is network caching in which a popular content is cached at the edge of the network in order to decrease the repetitive downloads. In this paper, a non-Cooperative Hierarchical Reactive Caching (nCHRC) algorithm is proposed as an optimal multi-level cache content placement method. The main innovative idea of this work is using a parameter called virality which illustrates the impact of time on the popularity of the contents. Moreover, some parameters such as well-known popularity, virality and size of the file are efficiently aggregated as a single parameter named priority. The results show that our method successfully improves the cache hit rate, backhaul traffic and content access delay of the network.