Retrieval and organizing information - Retrieval Information Models

Models based on diffuse logic

As the fact of having to make an initial estimate of weights in models likely to be more complex and therefore less striking for its use, the model of diffuse logic does not assign weight (in this case called degree of membership) initially.

If we continue comparing the model with the probabilistic model based on diffuse logic, the result of the latest probability calculation brings us back to the model if and when terms are used in document relevant documents and irrelevant. But in the model of diffuse logic, when the equation instead of the calculation as in the probabilistic model, now is defined taking into account the degree of membership of the terms. Therefore in this nearly whether the degree of membership is high increases the possibility that this term is in the document with a greater degree of relevance.

An additional advantage to this model, the use of diffuse models is highly recommended to solve problems of incompleteness and inaccuracy in the indexing of documents.

This page has been developed for one Computer Engineering subject of Carlos III University of Madrid, specifically, Recovery and Access of Information.

Versions available:

Download the DOC article