Discovering malicious locations in a public computer network

摘要

A web threat protection system may receive candidate uniform resource locators (URLs) from several URL sources. The candidate URLs may be received in a submission database. At least a portion of the candidate URLs is selected for further investigation by sending crawlers to retrieve objects from the selected URLs. The retrieved objects may be analyzed to determine whether they are malicious or good (i.e., not malicious). The result of the analysis may be used to build a security states database that includes security information of the selected URLs. Good URLs may be included in a safe URL sphere, which may be used to navigate to good websites on the Internet.