Back to search

National Technical University of Athens (NTUA)

Διαχείριση μεταβολών περιεχομένου ιστοσελίδων ιστού

Abstract

dc:description

This thesis has four main parts: • Part I (i.e., chapters 2 and 3) introduces a broad background of Web platform that includes HTTP protocol, markup languages and Web content types. We classified Web content into three types: (l) Data with no context, represented in plain text files (2) Information with a relationship between data, which depends largely on context for its meaning (3) Knowledge that is self- contextualizing. Based on this, we discussed current search engines and classified them into three generations: text-based, information-based and knowledge-based search engines. Text-based search engines have no state of art in their searching algorithms. Information-based search engines analyze the link structure at Web-level and use latent semantic indexing at document level. Semantic Web standards (e.g., XML, RDF, and OIL+DAML) are used as key drivers for knowledge-based search engines. • Part II (Chapters 4 and 5) handles the problem of content freshness and update at the search engine side. To keep search engines current, they should be notified about the behaviour of the Web pages: how are they created, updated, deleted and so on. To achieve this, we monitor Web content from time and meaningfulness perspectives. For the time perspective, we proposed Meta updates to create a kind of coordination between a Web site and a search engine. We evaluated experimentally the proposed approach by comparing the results of crawling the same set of Web pages in two modes (1) Crawling with Meta updates (2) Typical crawling. The experiments showed that supporting a crawler Meta updates increases the freshness average by 45%. The approach has a trade-off between the cost of crawling and the degree of freshness for pages with high change frequency. For such cases we propose pushing updates via the “Web Monitor” tool that we have developed. The tool monitors the changes of pages and alerts the user whether to submit the current update to search engine(s) or not. As for the meaningfulness perspective, we define the importance of a change in terms of three factors. (1) The syntactical similarity between a page and its successive version, which views pages as set of shingles. (2) Semantical similarity by using Latent Semantic Indexing that represents pages as vectors and measures how a page is semantically close to its modified version. The experiments showed that Latent Semantic Indexing is closer to the human judgement of detecting meaningful updates. (3) The importance of the page itself in terms of its location in the publication tree and its hit frequency by users. Web Monitor tool was supported with the diff tools (e.g., LSIdiff, Location, hit frequency and syntactical diff). • Part III (Chapter 6) defines two concepts in managing Web content at managerial level (1) A methodology with itemized phases of content development and maintenance (2) Representing the know-how best practices in an accessible form to ensure interoperability. We serialized some standards of Semantic Web (e.g., RDF and OIL) to achieve this requirement. • Part IV (Chapter 7) models a framework for Web content management systems in terms of a set of agents (e.g., Monitor, Diff, Notify, and Log).

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2002

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Hattab, Ezzeddin

Subjects

dc:subject × 20

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/16658

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Hattab, Ezzeddin. Διαχείριση μεταβολών περιεχομένου ιστοσελίδων ιστού. National Technical University of Athens (NTUA), 2002. http://hdl.handle.net/10442/hedi/16658