Heterogeneous Data Clustering Considering Multiple User-provided Constraints

Clustering on heterogeneous networks which consist of multi-typed objects and links has proved to be a useful technique in many scenarios. Although numerous clustering methods have achieved remarkable success, current clustering methods for heterogeneous networks tend to consider only internal infor...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:International journal of computers, communications & control Ročník 14; číslo 2; s. 170 - 182
Hlavný autor: Huang, Yue
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Oradea Agora University of Oradea 01.04.2019
Predmet:
ISSN:1841-9836, 1841-9844
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Clustering on heterogeneous networks which consist of multi-typed objects and links has proved to be a useful technique in many scenarios. Although numerous clustering methods have achieved remarkable success, current clustering methods for heterogeneous networks tend to consider only internal information of the dataset. In order to utilize background domain knowledge, we propose a general framework for clustering heterogeneous data considering multiple user-provided constrains. Specifically, we summarize that three types of manual constraints on the object can be used to guide the clustering process. Then we propose the User- HeteClus algorithm to solve the key issues in the case of star-structure heterogeneous data, which incorporating the user constraint into similarity measurement between central objects. Experiments on a real-world dataset show the effectiveness of the proposed algorithm.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1841-9836
1841-9844
DOI:10.15837/ijccc.2019.2.3419