Marginal Median SOM for Document Retrieval

What is it about?

The self-organizing map algorithm has been used successfully in document organization. We now propose using the same algorithm for document retrieval. Moreover, we test the performance of the self-organizing map by replacing the linear Least Mean Squares adaptation rule with the marginal median. We present two implementations of the latter variant of the self-organizing map by either quantizing the real valued feature vectors to integer valued ones or not. Experiments performed using both implementations demonstrate a superior performance against the self-organizing map based method in terms of the number of training iterations needed so that the mean square error, or average distortion, drops to the 1/e=36.788% of its initial value. Furthermore, the performance of a document organization and retrieval system employing the self-organizing map architecture and its variant is assessed using the average recall–precision curves evaluated on two corpora; the first comprises of manually selected web pages over the Internet having touristic content and the second one is the Reuters21578, Distribution 1.0.

Featured Image

Why is it important?

A superior performance of the proposed variant with respect to the MSE curve related to the training phase of the algorithm, and the average recall–precision curve related to the retrieval effectiveness during the test phase has been demonstrated, when the basic SOM algorithm is replaced by the proposed MMSOM for document organization and retrieval.