Skip to main content
select distinct n.content as name, r.content as role, t.nameID as nameID, n.unitID as unitID, n.departmentID as departmentID, n.facultyID as facultyID, n.provinceID as provinceID, n.countryID as countryID from transaction as t, Name as n, Role as r where t.itemID='92369' and t.nameID=n.id and t.roleID=r.id
scopus



Low-cost parallel Text retrieval using PC-cluster

Researcher

เครือข่ายคณะผู้วิจัย


เครือข่ายนักวิจัย+ผลงานวิจัย (full screen)

Abstract

We present a parallel vector space based Text retrieval prototype implemented on a low-cost PC cluster running Linux operating system, using the PVM Message passing library. We also embed the inverted file structure into our proposed prototype for fast retrieval. From several experiments derived from the standard TREC-9 collection, this prototype can index up to 500,000 web pages per hour using a simple x86 machine. We also obtain 5.4 seconds Query response time on searching in the one and a half million TREC-9 web pages, using 2 machines. © Springer-Verlag Berlin Heidelberg 2001.

Query response time (2 items found) | Message passing (35 items found) | Text retrieval (4 items found) | Computer operating systems | LINUX- operating system | Information retrieval | Parallel vectors | Fast retrievals | Inverted files | Parallel text | Vector spaces | Space-based | Websites |

ต้นฉบับข้อมูล : scopus