At present, social networks are strongly used for personal or professional purposes. The social networks have accumulated a large amount of information, which can be used in different ways. One of such possible purposes is advertising. In particular, when a company pays for advertising in a social network, it wants to get an ideal deployment of the ads on the network. The ideal behavior aims at reducing the high costs to pay, the number of hours to design advertising, the difficulty in the creation of ads by lots, among other things. In particular, the optimization of the production of advertising on social networks, based on the performance of the ads, is not an easy task. If we make an incorrect design, this can result in losses for a company. In this sense, this paper presents an intelligent system of management of social network advertising, based on data mining techniques, to automatically produce ads. In addition, we test our adaptive mechanism of automated production (generation) of online advertising on Facebook. Our system carries out automatic modifications and improvements of ads. Facebook is one of the more popular social networks, being widely used by companies for advertising, in order to create contents attracting the attention of users.