Reference Extract is envisioned as a web search engine. Reference Extracts will be built for maximum credibility by relying on the expertise and credibility judgments of librarians from around the globe. Users get results weighted towards sites most often referred to by librarians at 1,400 libraries worldwide.