Basically, text-mining is a software-based approach to turn text, retrieved e.g. from online sources, into data for analysis. #Social Innovation (SI) research is often occupied with the identification of best practices, success stories, and ways to upscale social innovation endeavours. Traditional approaches in #Social Science research had to invest considerable resources to collect data on SI. Text-mining tackles this issue by using machine-based learning systems to process a vast number of sources on the internet. This enables the creation of research tools that allow users to systematically browse and analyse structured and categorised data on SI. Different sources relevant to SI, like e.g. the social media presence of practitioners or researchers alike as well as scholarly publications or web pages are made more accessible for SI research.
However, a tool based on machine learning can only deliver good results if it is trained accordingly. The H2020-funded KNOWMAK project is developing an online user interface where users can explore a mapping of Social Innovation projects. This user interface is fed by the above-mentioned machine learning engine, based on an ontology of SI and an extensive annotation process where SI researchers are teaching the machine “what to look for”. Grounded on the data annotations by experts, the interface can give hints to identify innovative practices remotely and more easily. For SI research in general, this approach brings interesting opportunities, tapping new dimensions of data on SI.
If you want to follow up on the latest developments in this area check the following sources: