Japanese Internet portal-site www.goo.ne.jp powered by InfoBee technology

摘要

The biggest Internet portal-site in Japan is NTT's "goo" site, www.goo.ne.jp. Its speciality is its ability to handle the Japanese language. It uses the most powerful information search engine for the Japanese language so it provides a comfortable environment for the users who want or who must read Japanese information. Unlike the manually structured directory style of "Yahoo!", the computer-based robot-style search mechanism in "goo" automatically retrieves information from the net. Furthermore, "goo" provides the most up-to-date information through its original service "mogitate goo" which means "freshly-squeezed goo". Information categories such as the news site "Nikkei goo", also work perfectly to help the user find what he is looking for. We describe the key technologies behind the "goo" site and its services. The keys are the "InfoBee" information handling technologies, such as the information retriever, crawler, indexer and Japanese analyzer. The system described doesn't require a high performance super-computer. It works effectively on a workstation-cluster, even a whitebox-based personal computer cluster. It is economical, scalable, and offers the same performance as a super-computer.