Integrate Full-Text Retrieval with Digital Archives System

1
Integrate Full-Text Retrieval with Digital
Archives System

• ReporterChia-Hao Lee
• Computer System and Communication Lab,
• Academia Sinica Institute of Information Science
• 2003.11.06

2
Outline

• Current full-text retrieval system
• What needs ?
• What problems in current DB retrieval ?
• How to integrate ?

3
Full-text Retrieval - System Structure
Crawler
MediaParser
Language
Data Source Modules
Detector
Indexer (IE)
Language
DB
Translator
Searcher (ISAPI ASP)
Language
Keyword
Fulltext
In
-
Class
Translator
Searcher
Searcher
Searcher
4
Full-text Retrieval - Platform

• MS Windows
• Web Server
• MS IIS (ASP)
• The part of core
• C Language (.dll , .exe)
• Database
• MS SQL Server

5
What Needs ?

• Web-pages retrieval
• Some problems
• Pages generated from PHP, flash
• Database retrieval
• Management Interface
• Multi-fields retrieval in the same table
• Miss character retrieval

6
What Problems in Current DB Retrieval ? I

• Current management interface
• Select / Insert Server
• Select / Insert Database
• Select / Insert Table Key-Field (primary key)
• What problems ?
• Need an account for the access of all databases
  in the server
• No delete function
• Key-Field must be INTEGER type

7
What Problems in Current DB Retrieval ? II

• Other problems
• As the indexed status of the table
• Keep the last indexed key value (primary key
  value)
• No support miss character retrieval
• The last presentation after search result
• Limits
• The maximum of multi-fields retrieval is 5 in the
  same table
• The maximum retrieval length of a field is 1xxx
  characters

8
How to Integrate?

• Current retrieval platform
• Windows?IIS?ASP?C(MFC)?MS SQL Server
• Solution
• Keep the current retrieval platform
• Translate Window platform to Linux platform

9

• Q A
• Thanks for Your Attention !!