In Czech: Klasifikace Suffix Tree fr├ízemi - srovn├ín├ş s metodou Itemsets

In Czech: Klasifikace Suffix Tree fr├ízemi - srovn├ín├ş s metodou Itemsets

Classification based on Suffix Tree Phrases in Comparison with the Itemsets method

In this paper we present a text classification method using Suffix Tree (ST) phrases. We describe how to obtain ST-phrases from the training corpora, how to evaluate them and use them for text classification. Advantages and disadvantages of this approachare discussed and compared to the Itemsets method, which the Suffix Tree classification is based on. We also explain the way a threshold for multiclassclassification is determined. We devote some time to examine the document length influence on classification effectiveness and also compare the impact of higher order Itemsets and ST-phrases in both methods. Of course, some comparison of the results obtained with other favourite text classification methods is provided at last.

Keywords: text classification, document collection, itemsets, Suffix Tree, document evaluation, threshold determination

Year: 2005

Download: download Full text [301 kB]

Authors of this publication:

Roman Tesa┼Ö

Phone: +420 377632479

Roman is a PhD student at the Department of Computer Science and Engineering, Faculty of Applied Sciences, University of West Bohemia in Pilsen, Czech Republic. His work is focused on the utilization of word n-grams in text classification and document filtering.

Karel Je┼żek

Phone:  +420 377632475

Karel is the former group coordinator and a supervisor of PhD students working at research projects of this Group.

Related Projects:


Internet Content Filtering

Authors:  Roman Tesa┼Ö, Karel Je┼żek
Desc.:This project includes Web sites processing, analyzing, classification by means of their content and searching for other Web sites with similar content.

Document Classification

Authors:  Ji┼Ö├ş Hynek, Karel Je┼żek, Michal Toman, Roman Tesa┼Ö, Zden─Ťk ─îe┼íka, Petr Grolmus
Desc.:Use of inductive machine learning methods in classification of short text documents.