Local and global topics in text modeling of web pages nested in web sites

Topic models assert that documents are distributions over latent topics and words. A nested document collection has inside a higher order structure such as articles in journals, podcasts within authors, or web pages sites. In single of documents, global shared across all documents. For sites, topic frequencies likely vary sites site, almost certainly from page to page. hierarchical prior for this with distribution, site varying around the distribution. Web one United States local health department often contain geographic news not found on other some unique an individual site. Regular ignore nesting may identify but cannot label those nor corresponding owner. Explicitly modeling identifies owning local. US data, coverage is defined at level after removing words pages. Hierarchical can be used study how well covered.