Creator |
|
Language |
|
Publisher |
|
|
Date |
|
Source Title |
|
Vol |
|
Issue |
|
First Page |
|
Last Page |
|
Publication Type |
|
Access Rights |
|
Crossref DOI |
|
Related DOI |
|
|
Related URI |
|
|
Relation |
|
|
Abstract |
An optimal combination method is proposed to unify similarity matrices generated from relational graphs having various relations among scientific articles. Three relations are considered, that is, the... citation relation, the key word (or title) relation and the author relation. Canonical correlation analysis and alpha factor analysis are applied to determine the optimal combination coefficient. Maximal consistency and maximal generalizability are adopted as the criteria of optimization. The unified similarity matrix is analyzed by combinatorial cluster analysis. The optimal consistency model and its exploratory result are described.show more
|