Wednesday, 20. March 2013 20.03.13 09:34 Age: 5 Jahre

A Flood of Papers and Other Mysteries
A Research Network Brings Light into the Chaos of Scientific Research

The author network of scientist Mike Sharples in the mLearn conference series with a number of publications and co-author relations.

Stacks of paper pile up on any scientist’s desk – a sheer number topics and authors, publications that are relevant and many others that are not. Especially scientists who are at just starting their research career often have difficulties getting a general overview of their topic and keeping it: Which colleague is working on a similar topic? Whose paper should I read, who can I cooperate with, where can I ask for assistance? Now and then, even experienced researchers lose track of what they already published or what references they used before.

To support researchers in their scientific endeavors, Dr. Wolfgang Reinhardt from the Didactics of Computer Science research group led by Prof. Magenheim had the idea of developing a suitable research network. The goal of the network is to provide information on the people, their publications, projects, institutions and events, as well as their (temporal) relationships to each other – an interesting goal for a very complex system.

There are already a number of different networks in the Internet that could be used to reach these goals. On the one hand, however, they do not provide good opportunities for scientific use. On the other hand, scientists prefer to use those networks for private activities.

Together with the project group PUSHPIN, Dr. Reinhardt therefore developed a social network specifically by and for researchers. The goal is to create a comprehensive network of scientists through their publications.

The users themselves upload their publications into the system, so it constantly expands. Technically, this is done by a classic upload or by automatic synchronization with the users’ Mendeley account. Each paper is automatically analyzed as a digital version with its full text, and important information such as the title, the author, keywords, quotations and references are extracted. Particularly these references create the comprehensive research network of the authors and publications. Frequent or multiple reference entries show the entry’s degree of connection. With a slider function, users can filter for similar content or for referenced similar publications. In addition to downloading proposed papers, a number of further social interactions are also possible.

Besides PUSHPIN, Dr. Reinhardt has also developed ginkgo, a social conference management system with particular focus on scientific events. Both applications are being used to support scientific awareness and foster social interaction between researchers, and in principle they are conceivable in any research area.

So far, nearly 6,000 publications with 8,600 references and 250,000 authors from several conference series have been collected. The demand for further development is very high. In the future, further basic system functions are plausible. For example, work is currently being done on plagiarism detection and improved proposal algorithms.

Along with Dr. Reinhardt and the PUSHPIN project group, a number of successful final papers and papers by student assistant work in the research group are also involved in the project. The analyses of the papers are performed on the Hadoop Cluster for Large-Scale Analyses Publication (HCPA) and sponsored by 8,000 euro from the Commission for Research and Young Scientists.

PUSHPIN and ginkgo are available at the following Internet addresses:

pushpin.cs.upb.de
ginkgosem.com

Contact:
Dr. Wolfgang Reinhardt
Didactics of Computer Science
Tel: 05251 60-6603
wolle[at]uni-paderborn.de


Files:
FiS_November2012.pdf799 K