This study aims to analyze from the comments of the Chilean media how the use of social network sites (SNS) is conceived in the elderly. We used Latent Dirichlet Allocation (LDA) to generate a model that categorizes the media opinions about more senior people and SNS to analyze Twitter data.  The proposal was that there are stereotypes about the relationship between the elderly and SNS (for example, the elderly use SNS to be close to relatives, the elderly do not use SNS to start a business, among others) and that stereotypes are a cultural barrier to face the problem of heterogeneity in the use of these platforms. The data used in this study is twitter data taken is Twitter data from the official Twitter accounts of Chilean media between 2016 and 2020. In general, the steps of this study were 1) retrieve Twitter data using Twitter API; 2) data pre-processing, including basic cleaning, normalization, stemming, and stopword removal; 3) development of the LDA model, this step includes the search for the optimal number of topics; 4) interpretation of the model result. The result of the analysis LDA of the Chilean media tweets will be compared with a future survey on the motivations for using that we will carry out with Chilean elderly SNS users.