The search for information of real interest for a particular topic has always been a challenge, especially when the source of this information is on the Internet, where there are many digital media, platforms, databases and from which you can explore data. Large companies such as Google or Yahoo help us find these sources of information through their search algorithms, but they are not very effective when filtering a particular information, so it is necessary to formulate a model that responds to this need, in the present The document proposes a model based on 3 components, using information filtering tools, word valorization techniques based on their concurrences, weights, anti-words and a proposal to apply a word decay log within the information, in order to maintain a “score” and avoid false weights (triggered weights) of the information.