Nestriktās klasterizācijas algoritmi un to lietojumi tekstu apstrādē
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
lav
Abstract
Darbs ir veltīts klasiskās un nestriktās klasterizācijas algoritmiem, kuri tika pielietoti teksta klasterizācijai. Tiek apskatīti divi klasiskās klasterizācijas algoritmi, kas ir KM un hierarhiskās klasterizācijas metodes, un divi nestriktas klasterizācijas algoritmi: FCM un FCMdd klasterizācijas. Klasterizācijas algoritmi tika izmantoti “BBC News” ziņu portāla rakstu klassifikācijai pa piecām tēmām: bizness, tehnoloģijas, izklaide, sports un politika. Tika analizēti un interpretēti skaitliskie rezultāti.
This paper is devoted to classic and fuzzy clustering algorithms, which are applied to textual data in order to classify it into groups. Paper contains description of classic clustering algorithms such as KM clustering algorithm and hierarchical clustering algorithm, as well as fuzzy clustering algorithms: FCM clustering algorithm and FCMdd clustering algorithm. Clustering algorithms have been applied to “BBC News" texts in order to cluster them in five categories such as business, technology, entertainment, sports and politics. Obtained numerical results are analyzed and interpreted.
This paper is devoted to classic and fuzzy clustering algorithms, which are applied to textual data in order to classify it into groups. Paper contains description of classic clustering algorithms such as KM clustering algorithm and hierarchical clustering algorithm, as well as fuzzy clustering algorithms: FCM clustering algorithm and FCMdd clustering algorithm. Clustering algorithms have been applied to “BBC News" texts in order to cluster them in five categories such as business, technology, entertainment, sports and politics. Obtained numerical results are analyzed and interpreted.