In Internet of Things (IoT), numerous devices are able to collect and report data while they can execute simple processing tasks to produce knowledge. IoT Nodes exhibit limited computational re-sources, thus, they can only perform a limited number of tasks and store a short version of the collected data. In this paper, we propose a scheme that focuses on a distributed model for data storage in a group of IoT nodes. Nodes cooperate each other exchanging statistical information for their data. Our work aims to provide a model for the selection of the node where the incoming data should be stored irrelevantly of the node in which they are initially reported. The selection process involves a decision making process that adopts a statistical similarity model of the incoming data with the datasets present in the group, the estimation of the load of each node and the in-network communication cost. All these parameters are fed into a multi-class classification scheme for the final decision. Our aim is to have a view on the statistics of the available datasets beforehand, thus, facilitating the post-processing and the production of knowledge. We report on the evaluation of our scheme and present experimental results towards the presentation of pros and cons of our model.