|
Title:
|
DOCUMENT RETRIEVAL USING FUZZY MODELING OF NEURAL NETWORK |
|
Author(s):
|
Habib Karbasian , Siavash Kayal |
|
ISBN:
|
978-972-8924-56-0 |
|
Editors:
|
Nuno Guimarães and Pedro Isaías |
|
Year:
|
2008 |
|
Edition:
|
Single |
|
Keywords:
|
Adaptive neuro-fuzzy inference engine (ANFIS), document retrieval, neural network, vector space. |
|
Type:
|
Short Paper |
|
First Page:
|
363 |
|
Last Page:
|
367 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Document retrieval science is becoming a major field in computer science according to increase in the size of collection
of documents the user wants to search in. The initial purpose of this study is to retrieve documents using neural network
as a categorization tool that clusters documents in certain topics then the output of neural network is treated as a vector
and the distance of it is computed along with each documents desired output. Two models of neural network have been
utilized; the first model that has several nodes in the output representing topics and the second neural network has a
number of nodes which are associated with documents. We have found out that the first model works better than the
second. The advantage of conceptual search engine over the other kind of search engines based on exact matching is more
precise in retrieving documents. Then an improvement has been made with fuzzy modeling of the selected neural network
paradigm. Finally we propose that fuzzy-modeled neural network is more dominant than neural network itself. |
|
|
|
|
|
|
