Finding posts in digital libraries of authors with garbled names

Authors with names containing characters outside the ASCII character set have often names garbled also in digital libraries. We solve this problem using professional social networks and digital libraries relations, links and metadata. To find publications we use simple analysis of connections of the author, his co-authors and affiliates and an approximate string matching algorithm for estimating whether the garbled name can be the right name or not. Experimental results are very promising based on real data in digital libraries such are IEEE Xplore, ACM DL and SpringerLink. Description and explanation of the method is contained, a case study was used for Czech and Slovak authors.