Transkribus. A Platform for Automated Text Recognition and Searching of Historical Documents

Abstract

The Transkribus platform provides services and tools for the digitization, transcription, recognition and searching of historical documents. It is the only platform worldwide were non-technical users are enabled to train their own machine learning based neural networks and to apply them on their documents in order to generate an automated transcription and to make them searchable via keyword spotting. Transkribus is used by thousands of users and hundreds of archives, libraries, and research groups all over the world. In this paper we briefly describe the approach of the platform in terms of the underlying business and governance model, as well as the technical aspects of the platform.