Monday, March 22, 2010

Quo Lashes Buy Online

Information Retrieval Models for Information Retrieval.

Design SRI a is performed under a model, which is defined as "how to obtain the representations of documents and consultation, the strategy for assessing the relevance of a document concerning a consultation and methods to establish the importance (order) of the output documents "(Villena Román, 1997). There are several proposed classifications of models, one of the most comprehensive synthesis is performed Dominich into five groups (Dominich, 2000):

Baeza-Yates
classifies information retrieval models based on the initial task performed by the user the system: (1) retrieve information through an equation of search (retrieval) inserted in a form designed for this, or (2) taking the time to see (browse) documents in the search for references (Baeza-Yates, 1999), ushering in his classification hypertext [CON 1988] [NIE, 1990] model which is based web [BER, 1992].

This author divides the recovery-based models into two groups: traditional and structured. In the first of these includes the Boolean models, vector space and probabilistic. Subsequently, it presents a series of alternative paradigms to each model: set theory (Boolean fuzzy sets and extended), algebraic (vector generalized latent semantic indexing and neural networks), and finally, probabilistic (inference networks and knowledge networks), the structured models correspond to non-overlapping lists of terms and nearby nodes (they are sparsely distributed models). Models based on navigation between web pages are of three types: flat structure, structure and hypertext tour.

The first is a simple reading of a document in isolation from the context, the second has the potential to facilitate exploration by organizing the documents in a directory-type structure with a hierarchy of classes and subclasses and the third is based on the idea of \u200b\u200ba system information that the possibility of acquiring information in a sequential but not strictly through nodes and links (Baeza-Yates, 1999). Baeza-Yates is also the provider of an additional classification of these models information retrieval, performed according to the method of consultation and the logical view of documents:



BIBLIOGRAPHY:

Both Baeza-Yates ( Baeza-Yates, 1999) and Villena Román (Villena Román, 1997) conducted a detailed presentation of each of the models, and also interesting reading Grossman and Frieder [GRO, 1998], to discuss alternatives to the models classics.

0 comments:

Post a Comment