Welcome!
To use the personalized features of this site, please log in or register.
If you have forgotten your username or password, we can help.
My Menu
Saved Items

Optimization of Restricted Searches in Web Directories Using Hybrid Data Structures

Fidel CachedaContact Information, Victor CarneiroContact Information, Carmen GuerreroContact Information and Angel ViñaContact Information

(5)  Department of Information and Communications Technologies, Facultad de Informática, Campus de Elviña s/n, 15.071 A Coruña, Spain
Abstract
The need of efficient tools in order to manage, retrieve and filter the information in the WWW is clear. Web directories are taxonomies for the classification of Web documents. These kind of information retrieval systems present a specific type of search where the document collection is restricted to one area of the category graph. This paper introduces a specific data architecture for Web directories that improves the performance of restricted searches. That architecture is based on a hybrid data structure composed of an inverted file with multiple embedded signature files. Two variants are presented: hybrid architecture with total information and with partial information. This architecture has been analyzed by means of developing both variants to be compared with a basic model. The performance of the restricted queries was clearly improved, especially the hybrid model with partial information, which yielded a positive response under any load of the search system.1
This work has been partially sponsored by the Spanish CICYT (TIC2001-0547).

Contact Information Fidel Cacheda
Email: fidel@udc.es

Contact Information Victor Carneiro
Email: vicar@udc.es

Contact Information Carmen Guerrero
Email: clopez@udc.es

Contact Information Angel Viña
Email: avc@udc.es
Fulltext Preview (Small, Large)
Image of the first page of the fulltext

References secured to subscribers.



Export this chapter
Export this chapter as RIS | Text
 
Remote Address: 38.107.191.108 • Server: mpweb03
HTTP User Agent: CCBot/1.0 (+http://www.commoncrawl.org/bot.html)