Fast Forward Index Methods for Pseudo-Relevance Feedback Retrieval

Abstract

The inverted index is the dominant indexing method in information retrieval systems. It enables fast return of the list of all documents containing a given query term. However, for retrieval schemes involving query expansion, as in pseudo-relevance feedback (PRF), the retrieval time based on an inverted index increases linearly with the number of expansion terms. In this regard, we have examined the use of a forward index, which consists of the mapping of each document to its constituent terms. We propose a novel forward indexbased reranking scheme to shorten the PRF retrieval time. In our method, a first retrieval of the original query is performed using an inverted index, and then a forward index is employed for the PRF part. We have studied several new forward indexes, including using a novel spstring data structure and the weighted variable bit-block compression (wvbc) signature. With modern hardware such as solid-state drives (SSDs) and sufficiently large main memory, forward index methods are particularly promising. We find that with the whole index stored in main memory, PRF retrieval using a spstring or wvbc forward index excels in time efficiency over an inverted index, being able to obtain the same levels of performance measures at shorter times. . 2015. Fast-forward index methods for pseudo-relevance feedback retrieval.

A Principled Approach Using Fuzzy Set Theory for Passage-Based Document Retrieval

“…A second retrieval, i.e. a re-retrieval [58] is performed by calculating new scores for all passages in the collection with the expanded query Q P RF . In our experiments, the parameters α m and N QE are determined by calibration to yield the highest performance metric.…”

Comparative Analysis of Ranking Functions for Retrieving Information From Medical Repository

“…Search evaluation is user-centered, and users are sometimes uncertain and ignorant of the content they are looking. A user Interaction [5] or user Feedback [6] [7] can enable the IR system to identify the content. The query is essential in understanding user intent.…”

“…The Boolean similarity measures are Jaccard Dissimilarity, Dice Dissimilarity, Canberra Distance and their formulas as presented in (5) and (6).…”

“…Query representation: Weights are assigned, to evaluate the output argument in the range of 0 to 1. Similarity operators: Extended / Soft Boolean Model Pnorm proposed by Salton [3][4][5][6][7] based on similarity correlation formulae ORED with Anick Approaches where,…”