The Design of Text Signatures for Text Retrieval Systems

1995

This paper describes the results of a preliminary attempt to use a genetic algorithm to divide an inverted file into a specified number of partitions such that the total number of documents indexed by a particular partition is approximately equal to the total number of documents indexed by each of the other partitions. The purpose of identifying such equifrequent partitions is to assist in the generation of text signature representations of documents which are more discriminating than those created using more traditional techniques. The paper is divided into six sections. Following the introduction, the second of these describes the main idea behind the signature approach The third section introduces the idea of genetic algorithms and briefly reviews earlier work on the application of this technique to information retrieval problems. The fourth section describes how we have used a genetic algorithm to partition a section of the inverted file used to index the LISA document test coll...

1998

Hypertext systems provide an appealing mechanism for informally browsing databases by traversing selectable links. However, in many fact finding situations string search is an effective complement to browsing. This paper describes the application of the signature file method to achieve rapid and convenient string search in small personal computer hypertext environments. The method has been implemented in a prototype, as well as in a commercial product.

Computer Science and EE Dept., University of …, 1995

A problem conamonly faced by users of the World-Wide Web (WWW) is forgetting the path traversed to reach a previously read document. SWISS (Seeking World-Wide Web Information Using a Signature File Search) is a system designed to alleviate this' lost ...

2003

This paper presents a new approach for document image decomposition and retrieval based on connected component analysis and geometric properties of the labelled regions. The database contains document images with Arabic/Persian text combined with English text, headlines, ruling lines, trade mark and signature. In particular, Arabic/Persian signature extraction is investigated using special characteristics of the signature that is fairly different from English signatures. A set of efficient, invariant and compact features is extracted for validation purposes using angularradial partitioning of the signature region. Experimental results show the robustness of the proposed method.

1991

This paper discusses the implementation of nearest-neighbour document retrieval in serial files using transputer networks. The system uses a two-stage retrieval algorithm in which an initial text-signature search is used to exclude large numbers of documents from the detailed and time-consuming pattern-matching search. The latter is implemented using a processor farm, so that documents which match at the signature level can be examined in parallel to determine whether they are, in fact, a good match for the query. The results demonstrate that communication is the critical factor in all of the transputer networks that were investigated. A high degree of speed-up can be obtained when only the pattern-matching search is carried out. When text signatures are used, however, the speed-up is less, decreasing in line with an increase in the size of the text signatures that are used.

2007

The purpose of this research is to give an idea about inverted files and signature files based on Arabic documents collection, and to give the comparison points between the two techniques and the performance of the two techniques on each of the comparison points. The most common measures of system performance used to compare the information retrieval mechanisms are time, space, and recall/precision evaluation measurements. The shorter the response time is, the smaller the space used, the better system is considered to be [1], so our comparisons point will include space overhead, search time, and average recall/precision . In this research, two indices will be built, inverted-file and signature-file. However, to measure the performance of each one, a retrieval system must be built to compare the results of using these indices. A collection of 242 Arabic Abstracts from the proceeding of the Saudi Arabian National Computer Conferences have been used in the two systems, and a collection...

Journal of Discrete Algorithms, 2013

Signature file is a well-studied method in information retrieval for indexing large text databases. Because of the small index size in this method, it is a good candidate for environments where memory is scarce. This small index size, however, comes at the cost of high false positive error rate. In this paper we address the problem of high false positive error rate of signature files by introducing COCA filters, a new variation of Bloom filters which exploits the co-occurrence probability of words in documents to reduce the false positive error. We show experimentally that by using this technique in real document collections we can reduce the false positive error by up to 21 times, for the same index size. It is also shown that in some extreme cases this technique is even able to completely eliminate the false positive error. COCA filters can be considered as a good replacement for Bloom filters wherever the co-occurrence of any two members of the universe is identifiable.

Lecture Notes in Computer Science, 1999

In this paper, a new indexing technique to support the query evaluation in document databases is proposed. The key idea of the method is the combination of the technique of pat-trees with signature files. While the signature files are built to expedite the traversal of object hierarchies, the pat-trees are constructed to speed up both the signature file searching and the text scanning. In this way, high performance can be achieved.

Proceedings ADBIS, 1997

Today, two classes of indexing methods enjoying wide applicability are the Inverted Index and the Superimposed Coding based Signature File (SC-SF). The former is most efficient in query processing but utilizes extra storage of size comparable to that of the textbase, whereas the latter is most efficient in storage utilization. The present study builds upon the results obtained in previous research [2], and proposes a hybrid structure for text retrieval. The new structure is labelled S-Index and is shown to be of a tunable ...

Information Sciences, 1994

A variation of the superimposed coding version of the signature file method is introduced. The processing of a user query is similar to that of the conventional method with one difference: the output is provided in a ordered (ranked) way. The latter is based on a credibility value reflecting the probability of each candidate block to pass the validation test rather than being a false drop. The principles, an analytical model, and simulation results of the new technique are presented. The model's usefulness is the reduced I/O activity for cases where the user is only sampling the text base and is satisfied by retrieving a limited number of the many documents that qualify.