DITHENS
The research project aims at developing an innovative tool for students and researchers, thanks to deep learning technology in the field of optical text recognition (OHR/OTR), providing automatic transcriptions of handwritten news-sheets from different cultural settings.
The staff at Pegaso offers in-depth competence in the field of early modern handwritten journals and in the early stages of several European languages, and a specific awareness concerning the main issues faced in the development of such tools. Co-operation with international agents will extend the research field to encompass the main European traditions.
We need partner related to the following topic:
- Firstly, the new software (based on algorithms such as Long Short-Term Memory and Multi-Dimensional Recurrent Neural Networks), created in co-operation with private or public agents, will be able to scan and read handwritten documents in early modern calligraphy, accounting for spelling variations and styles of personal handwriting.
- Secondly, a text comparison tool based on AI will detect textual borrowings or even plagiarisms from handwritten news-sheets by contemporary or later historians.
- Interdisciplinarity is the cornerstone of the project. The joint effort of historians and AI experts shall yield multifaceted insights concerning the circulation of news and information, which still today is one of the main features of European identity and values.
If interested, please contact giampiero.brunelli@unipegaso.it