Internet economic news gathering and classification: a neural network software agent based approach

The explosive growth on the amount of information available on the Internet makes it a hard task to select what is worth reading in our scarce available time. The project described in this paper tackles this problem using techniques taken from the areas of autonomous agents and artificial neural networks. An agent for economic news gathering and classification was designed implemented and successfully tested over sites about economy available on the Internet. Inputs to the system are news text picked up by software agents from selected Internet economic sites, and the outputs are those same news classified by topics of interest in at most six classes. The classification is based on an artificial neural network that runs inside the classification agent, especially trained with patterns that allow carrying out the desired news analysis and classifications.