Abstract:

The mismatch between the web advertisement and the current page leads to worse advertisement effect. In the paper,we use the Bayesian model extension based on site and the semantic extension based on Wikipedia to accurately extract web tag information.We use more accurate tag to match the network advertisement,enhancing the advertisement effect.This paper designs a web tag recommendation system base on semantic extension.The experiments confirm that the effect is good.

Key words: tag extraction;semantic extension;tag recommendation