Určení tematické konzistence dokumentu
| Title in English | Determining topic consistency of a document |
|---|---|
| Authors | |
| Year of publication | 2011 |
| Type | Article in Proceedings |
| Conference | Znalosti 2011 |
| MU Faculty or unit | |
| Citation | |
| Web | http://znalosti.ics.upjs.sk |
| Field | Informatics |
| Keywords | fulltext search engine; topic consistency; backlinks |
| Description | The aim of this work is to design and implement a tool, which should be able to assign a score reflecting topic consistency of any web document written in the Czech language. This score is dedicated to be used for deciding whether the document's hyperlinks are appropriate for computing relevancy of referenced documents. In fact, it turns out that inconsistent documents should not be used. The presented algorithm uses both statistical and heuristic methods and has the precision about 93.5 % on the set of 200 test documents. |
| Related projects: |