DescriptionThis paper applies text mining and network approaches to ERSA conference papers to identify the emergence, shifts, and inter-relationships between topics in regional science and the location of researchers in the periods 1998-2006 and 2010-2012. Two important aspects will be analyzed, first, what topics have been presented and discussed at the ERSA conferences? How did topics evolve and change over the time period under investigation? These questions will be analyzed on the basis of the titles and abstracts of the papers. Text mining techniques such as stemmed word frequencies, frequency data on word pairs, term-document matrices (TDM), and term-frequency-inverse-document-frequency (TF-IDF) matrices are used to identify thematic components, their elements and the change in the topic accumulation over time. Preliminary findings suggest as a stable common 'core' of topics, while identifying topical elements which join it and others which fail to sustain themselves in later years. Secondly, who contributed to the ERSA conferences? At what frequency did colleagues contribute? With whom did they collaborate in form of joint papers? For this part of the analysis this paper identifies individuals through available email addresses. These questions will be analyzed on the one hand by methods of descriptive statistics, on the other hand by methods of social network analysis. In a final step of the analysis we will try to combine the two aspects mentioned above for selected actors and topics. With this combined analysis we try to find out who introduced new topics to the conferences, and through what networks they got diffused. The paper will demonstrate the usefulness of combining text mining tools with social network analysis to the understanding of knowledge evolution and knowledge transfer in social science.
|Period||26 Aug 2014 → 29 Aug 2014|
|Event title||54th European Congress of the Regional Science Association (ERSA)|
|Degree of Recognition||International|
Project: Research funding