YabTab automatically converts web pages to tables. There is tonnes of information on web: think of product listing pages, course catalogues, job postings, reports - and all of them are essentially tables. Product listing pages, for example, are tables with one row for each product, columns for product information like name, features, price, etc.
However all current scraping tools either require extensive configuration to extract such information or are domain specific (yet rarely works). The idea behind YabTab is to build a tool which can auto-extract such "tabular" information from all websites irrespective of their domain or underlying structure and technology. YabTab uses revolutionary Machine Learning techniques to recognize these patterns in any web page, a skill only humans were capable of so far.