چکیده:
With the increase of the volume of information and the progress in technology, the deficiency of traditional algorithms for fast information retrieval becomes more clear. When large volumes of data are to be handled, the use of neural network as an A rtificial Intelligent technique is a suitable method to increase the information retrieval speed. Neural networks present a suitable representation of knowledge in retrieval appl ications. The nodes in neural network, present the items of information retrieval such as keywords, authors,. . . and lioks are used for data transfer between layers. Finally, it results in a network output that is a retrieval document. In this article, the use of SOM neural network for data clustering is shown. A model of SOM neural network for a sample information retrieval from INIS database is impl emented.
خلاصه ماشینی:
When large volumes of data are to be handled, the use of neural network as an A rtificial Intelligent technique is a suitable method to increase the information retrieval speed.
Abstract - Keywords - Information, Text Retrieval, Neural Network, Self Organizing Map, Clustering Algorithm.
To calculate the SC(Q,Di) (Similarity Coefficient) for all documents, one of the effective strategies in information retrieval is to use neural network systems.
Using neural January / June, 2007 Iranian Journal of Information Science & Tec h nology, Volume 5, Nu mber 1 network in 1 nformation retrieval has some advantages such as: I .
When the requested data (keywords) do not exist in the document collection, we can use the neural network to retrieve the information proximity around the required information.
January / June, 2007 Iranian Journal of Information Science & Tech no iogy, Volume 5, Number 1 (View the image of this page) Figure I : The SOM learning algorithm.
Iranian Jo urnal of Information Science & Technology UOlunie 5, Number t Jan uary / June, 2007 (View the image of this page) Figure 2: The project's algorithm.
In this research, simple Gausslan neighboring function is used as follows: (View the image of this page) Iranian Journal of Information Science & Tech nology, Volume 5, Number t The value of # is reduced by increasing the learning time /.
January / June, 2007 Iranian Journal of Information Science A Technology, Volume 5, Number 1 Pi glotyges ond data (View the image of this page) The SOM network with 3 dimensional presentation of data is shown in Figure 7.