| Author | Naresh Srikalra |
| Call Number | AIT Thesis no.CS-95-16 |
| Subject(s) | World Wide Web (Information retrieval system)
|
| Note | A thesis submitted in partial fulfillment of the requirement for the degree of Master of
Engineering |
| Publisher | Asian Institute of Technology |
| Abstract | World-Wide Web (WWW or Web or W3) is an Internet-wide distributed hypertext system operating on a client/server basis. Due to the rapid growth in diversity of information, user community, and information volume; it is increasingly difficult to achieve effective information retrieval in the Web, especially to retrieve the hypertext documents having the same sets of keywords, but different meaning. This work incorporates the advantage of Neural Networks architecture to solve such problem. The problem can be divided into two sections, i.e., textual and non-textual information retrieval. A Kohonen's Feature Map or Self-Organizing Map is used as a tool to solve this problem with a modification in the input data representation. A new indexing scheme is proposed for textual as well as non-textual information retrieval. The former employs the idea of associative memory, to store information of the links between the keywords in the documents; while, the latter combines all sets of keywords associated with a particular non-textual information to represent it as a new input. An improvement in the performance of the new textual indexing scheme as compared to the traditional Kohonen's Feature Map method is also shown. |
| Year | 1995 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Vilas Wuwongse; |
| Examination Committee(s) | Stueart, Robert D.;Vatcharaporn Esichaikul; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1995 |