The goals of the project are

  • to develop a set of new bibliometric indicators to identify outstanding participants in geographic and regional clusters,
  • to model the impact of spatial and thematic proximity on successful cooperation,
  • to evaluate the model hypotheses using the HypTrails approach from machine learning
  • and to develop new methods for the extraction of suitable indicators for regional competency and interaction from websites and social media.


The project is divided into four subprojects:

  • SP1: Identification of thematically and regionally outstanding actors,
  • SP2: AuthorTrails: Co-Author-Analysis,
  • SP3: Determinants of the origin and success of cooperation in regional research clusters,
  • SP4: Determinants of the origin and success of cooperation in regional innovation clusters.

Identification of thematically and regionally outstanding actors

In bibliometrics, a variety of measures to evaluate the “height” of an author, a publication or a journal is commonly used. These values result in rankings, sometimes limited to specific thematic disciplines. The validity of such rankings is limited:

Just like identifying the 100 highest mountains does not give any information about the mountain world beyond Asia, these rankings may lead to knowledge about just a small subcommunity that is not representative for the observed data.

Based on this analogy and on established bibliometric indicators, the goal of this subproject is to identify the “peaks” in the world of science: The regional outstanding researchers, publications and institutions.

AuthorTrails: Co-Author-Analysis

The goal of this sub-project is to propose a set of hypotheses about the creation of research and innovation clusters. To compare these hypotheses for plausibility, the HypTrails approach will be advanced and transferred to publication graphs.

The HypTrails approach allows to compare a variety of hypotheses with respect to a specific dataset. One major challenge will be the representation of datasets and hypotheses in a similar way to be able to choose the best hypothesis for the origin of a specific dataset.

Determinants of the origin and success of cooperation in regional research clusters

This subproject develops procedures for the extraction and analysis of indicators for thematic and geographic competency and interaction. This methods will be based on information extracted from the web and social media and will allow to describe and predict the creation of new research and innovation clusters.

This subproject will use methods for classification of websites and link data and procedures developed by subproject 1.

Determinants of the origin and success of cooperation in regional innovation clusters

Regional innovation clusters are characterized by interaction and cooperation of various participants. Which attributes help to predict the development of local cooperation between public and private research? Which role plays the presence of regional research clusters for the creation of new innovation clusters in specific thematic fields? How similar are actors that are responsible for the creation of new local interaction between participants that were not connected before?

The goal of this subproject is a contribution to the finding of answers to these questions. These answers will be developed in close consultation with the other subprojects by connecting a variety of individual datasets dealing with, for example, topic-related publications or patents, citations and the whereabout of PhD students.