Comprehensive Analysis of H2020 Funding Participation based on LDA Topic Modeling and Robust Outlier Identification

Being able to participate in Horizon 2020 research projects is an outstanding opportunity for innovation and development. Regions can also utilize incorporated knowledge, equipment, added value etc., furthermore can increase their overall competitiveness and resiliency in the long run. Corresponding literature actively investigates the formed social network of project members, central players and possible factors that might contribute to a successful project submission. In the present study freely accessible project information are analysed in combination with regional attributes based on the Minimum Covariance Determinant method. Results are supplemented with results of LDA topic modeling of textual project descriptions. Regional distribution of main project topics are provided and robust outliers are identified with regards of financial funding that indicates major differences between member states connected before 2004 and after.