Browse Prior Art Database

Method for High Speed Full Text Search

IP.com Disclosure Number: IPCOM000013682D
Original Publication Date: 2000-Nov-01
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 55K

Publishing Venue

IBM

Abstract

Disclosed are methods to get the results very quickly in the full-text search system. This paper includes two method. The 1st method is about Boolean operations. The 2nd method is about the full-text index and the other type's data. The 1st method is described as follows: The 1st method is to minimize amount of information to be read when the Boolean search operations are done, because an effective document number for all terms is searched for in order. Fig. 1 shows the logic.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 65% of the total text.

Page 1 of 2

Method for High Speed Full Text Search

Disclosed are methods to get the results very quickly in the full-text search
system.

This paper includes two method.
The 1st method is about Boolean operations. The 2nd method is about the
full-text index and the other type's data.

The 1st method is described as follows:
The 1st method is to minimize amount of information to be read when the
Boolean search operations are done, because an effective document number for
all terms is searched for in order.

Fig. 1 shows the logic.

STA R T

R eset an effective docum ent num ber (N )

101

102

G et docum ent num bers not less than N , ( >=N ), in each term

103

104

NO

D o docum ent num bers not less than N exist in all term s?

YES

EN D

YES

A re the num bers sam e in all term s?

NO106105

Set the m aximum number among the next candidate docum ents to the term s as N

S tore the sam e num ber into the result set, and set the sam e num ber + 1 as N

Fig.1

An very simple example is shown to understand it easily.
query: PC and internet
number of all documents: 1,000,000
number of documents including PC: 1,000

1

[This page contains 7 pictures or other non-text objects]

Page 2 of 2

document numbers including PC: 1,000, 2,000, 3,000, ..., 989,000,
990,000, 991,000, .....
number of documents including internet: 2,000
document numbers including internet: 990,000, 990,005, 990,010, .....
index structure: the document number is in ascending order

1000 and 990000 are gotten at the 102 step in Fig.1.
990000 is set...