Detection of trends of internet traffic using sequential patterns/Interneto duomenu srauto tendenciju nustatymas naudojant tipines nuosekliasias sekas

We propose a new method of possible trend detection in the internet traffic by using sequential pattern analysis. The main steps of this method, such as the template pattern discovery in time--series data and pattern comparison and matching for sequential pattern analysis, are analyzed and the possi...

Full description

Saved in:
Bibliographic Details
Published inElektronika ir elektrotechnika no. 5 (93); p. 3
Main Authors Jelinskis, J, Lauks, G
Format Journal Article
LanguageEnglish
Published Kaunas University of Technology, Faculty of Telecommunications and Electronics 01.06.2009
Subjects
Methods
Object recognition (Computers)
Pattern recognition
Sequential analysis
Technology application
Latvia
Online AccessGet full text
ISSN1392-1215

Cover

More Information
Summary:We propose a new method of possible trend detection in the internet traffic by using sequential pattern analysis. The main steps of this method, such as the template pattern discovery in time--series data and pattern comparison and matching for sequential pattern analysis, are analyzed and the possible constraints and problems are pointed. Experimental data are depicted and future research subjects are described. Ill. 6, bibl. 9 (in English; summaries in English, Russian and Lithuanian). Siulomas naujas budas interneto duomenu srauto tendencijoms nustatyti naudojant tipines nuosekliasias sekas. Analizuojami pagrindiniai metodo etapai--tipines sekos aptikimas laikui begant kintanciuose duomenyse, ju palyginimas ir sutapatinimas atliekant nuosekliuju seku analize. Nurodomi galimi apribojimai ir potencialios problemos. Pateikti eksperimentiniai duomenys ir aprasomos ateities tyrimu kryptys. Il. 6, bibl. 9 (anglu kalba; santraukos anglu, rusu ir lietuviu k.).
ISSN:1392-1215