ICC-Crawler is operated by Japan's National Institute of Information and Communications Technology (NICT) Universal Communication Research Institute to collect web pages for natural language processing and AI research.