Adaptive text-data clustering by the semi-supervised learning

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62156489%3A43110%2F09%3A00147093" target="_blank" >RIV/62156489:43110/09:00147093 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Adaptive text-data clustering by the semi-supervised learning

Popis výsledku v původním jazyce

Many current real-world applications as, for example, Internet e-commerce, have to deal with huge volumes of predominantly textual data. The data include hidden information like potential categories of similar customers, suppliers, producers, etc. Thesecategories usually can change during the time. The paper discusses the possibilities of looking for clusters that represent individual classes. One of the main problems is the standard clustering methods give often unsatisfying results by unsupervised-learning methods. However, having very small initial subsets of good examples, the clustering can dramatically improve its results, providing much better clusters applicable to categorizing in the future. This method is known as the semi-supervised learning (SSL) from a limited number of examples. In the paper, some results of applying the SSL method to real-world unlabeled data instances are demonstrated and compared with selected traditional clustering algorithms. Using labeled examples

Název v anglickém jazyce

Adaptive text-data clustering by the semi-supervised learning

Popis výsledku anglicky

Many current real-world applications as, for example, Internet e-commerce, have to deal with huge volumes of predominantly textual data. The data include hidden information like potential categories of similar customers, suppliers, producers, etc. Thesecategories usually can change during the time. The paper discusses the possibilities of looking for clusters that represent individual classes. One of the main problems is the standard clustering methods give often unsatisfying results by unsupervised-learning methods. However, having very small initial subsets of good examples, the clustering can dramatically improve its results, providing much better clusters applicable to categorizing in the future. This method is known as the semi-supervised learning (SSL) from a limited number of examples. In the paper, some results of applying the SSL method to real-world unlabeled data instances are demonstrated and compared with selected traditional clustering algorithms. Using labeled examples

Druh

D - Stať ve sborníku

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

—

Návaznosti

Z - Vyzkumny zamer (s odkazem do CEZ)

Rok uplatnění

2009

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

ASIS 2009 -- Adaptívne siete v informačných systémoch

ISBN

978-80-8094-593-0

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

—

Název nakladatele

Nitra

Místo vydání

Univerzita Konštantína Filozofa v Nitre

Místo konání akce

Nitra

Datum konání akce

1. 1. 2009

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

—