Abstract

The search for information on the Internet is often not an easy job for many Internet users. Because of the lack of a standard indexing scheme and an informative query interface, a Net search could have thousands of hits returned and the number of search misses is high. This thesis is part of the work to develop an Indexing and Searching System for the Internet called the CINDI (Concordia INdexing and DIscovery) system. It is aimed at providing a standard index scheme called Semantic-Header and informative query interface for users and providers of resources published on the Internet. This thesis presents the architectural design of the CINDI system, the design and implementation of the client part of the expert system, and the design and implementation of the graphical user interface (GUI) for the CINDI system. As the interface is an important part for the quality of a software, it has been carefully designed and implemented in an effort to be easy to use, user friendly, and consistent. The interface has been implemented under the UNIX system using the Motif toolkit and C programming language. The heart of the indexing system is the record called Semantic-Header that is kept for each item being indexed. The grammar of the Semantic-Header and that of the search query are also discussed in this thesis. An interface between the CINDI system and Netscape Navigator is also implemented. Finally, some directions for future work related to CINDI are described.