The web-based CINDI (Concordia INdexing and DIscovery) System attempts to solve the common problem of the current search engines, such as lack of a standard indexing and an information query interface. It proposes Semantic Header as the standard index scheme, and stores the index entry into MySQL database management system. By invoking a set of graphic user interface, the CINDI system allows the resource contributor to catalog his own resource and enables the user to search for hypermedia documents based on the title, author(s), keyword(s) and subject search criteria.  This report describes the redesign and implementation of the Semantic Header database, and the design and implementation of the graphic user interface for the resource registration subsystem and resource search subsystem. The user-friendly interface has been implemented with PHP script language on the Linux platform. The three-tier client/server architecture is used in the web-based CINDI system, with Apache web server exchanging information between the web Browser (the client) and the backend MySQL database.