Indeed the internet contains tons of very helpful and versatile public domain information, but very often these pieces of interesting information are awfully distributed and presented in an inappropriate format to form a general answer to the question of interest.
The dwh GmbH has not only developed automatic web scraping methods for thousands of different urls and analysis tools for xml files, but also uses state-of-the-art knowledge from natural language processing to systematically obtain textual information.
Using this approach, the used time allotted to politicians for speaking in the Austrian parlament was analyzed. But also questions like ‘In which Austrian county are the most general practitioners per capita available on Friday afternoons?’ can be answered.