Publication Details
Analyzing Logical Structure of a Web Site
HTML analysis, Semi-structured data, Information extraction
The today's World Wide Web consists mainly of documents written in Hypertext Markup Language (HTML). This language has been developed for describing the look of the documents and the references to other documents and therefore it has very poor facilities for describing the semantics and the structure of the contained data. Moreover, some of these facilities are often not used by the authors of the documents or they are not used in apropriate way. In our work, we are attempting to analyze the look and the stucture of a Web site represented by the facilities of the HTML language and create its logical model which would represent the data relations the same way a human user would see it. We propose a tree representation of a Web site and algorithms for the analysis of the most importatnt HTML constructions - section headings, lists, tables and links.
@INPROCEEDINGS{FITPUB6920, author = "Radek Burget", title = "Analyzing Logical Structure of a Web Site", pages = "29--35", booktitle = "Proceedings of 5th International Conference ISM '02 - Information Systems Modelling", year = 2002, location = "Ostrava, CZ", ISBN = "80-85988-70-4", language = "english", url = "https://www.fit.vut.cz/research/publication/6920" }