This blueprint extracts out title, description and body from html either via xpath or by automatic cluster analysis - View it on GitHub