Text mining to identify skills, stakeholders and capabilities: the case of Artificial Intelligence in Emilia-Romagna

In 2020, SIRIS Academic presented a poster at the 6th World Open Innovation Conference.

The study presents a semantic analysis to map the science, technology and innovation (STI) activities in the field of Artificial Intelligence in the Emilia-Romagna region (Italy). The study aims at fostering research-to-business collaborations; detecting existing and potential capabilities to address challenge-driven innovations. 

We identified which topics are covered, what is the relative regional specialisation, which are the key actors, and which are the internal and external linkages. Finally, we measured how the topics have been evolving in time and how the specific domain of Artificial intelligence is applied to solve societal challenges (SDGs).

Methodology

We analysed single STI records proceeding from different repositories. Text mining techniques were applied to the publication abstracts, patent descriptions and R&I project objectives to extract a wealth of textual information describing in detail STI activities and results.

The methodology uses both automated text-mining computer techniques and qualitative and quantitative analysis.The approach is divided into the following steps:

1. Extraction of documents produced by the regional research ecosystem;
2. Definition of a vocabulary of key concepts that define the domain of research in Artificial intelligence
3. Use of the vocabulary to identify, among the texts extracted in step 1, all those related to the domain of interest
4. Fine-grained, unambiguous identification of the actors 
5. Quantitative analyses using the extracted data

‍