Implementing High Performance Retrieval Process by Max-Score Ranking
U. N. Vignesh · IOSR Journal of Computer Engineering · 2013
This paper presents a comparison report of two different processes of retrieving a keyword or data's from a given database or from a multiple databases.The process1 known as Extended Boolean Retrieval (EBR) model, it gives us an output from the database.Since EBR model implementation aspects lead to a high cost, we consider an p-norm approach to the EBR implementation.P-norm approach plays a role in the EBR model to maintain strictness of the conjunctions and disjunctions to set them with their own identification on the considerable node.The process2 known as Ensemble Learning Paradigm (ELP).In this paradigm of text categorization aspect, first it assigns a value to a given keyword or data and then starts it's searching process from an index.This value contains the factors such as a position and appearances of word.In existing, they use these concepts in Bag-of-word approach.In this paper EBR model gives an advantage of reformulation aspect, which gives a hundreds or thousands of answers for the given query.In ELP, term frequency identification paves the way to produce a result based on the frequencies of an regarded query in the database.To end, we evaluate with the reported results of these models on query to prove an better retrieving process based on their efficiency and accuracy with the max-score ranking algorithm.