Machine Learning

We are exploring how Machine Learning can be applied to archival descriptions and also digital content. This part of a project that will run through 2022-2023, Images and Machine Learning. 

Cloud services are being used which include built-in machine learning models have been used to experiment with (1) analysing the textual descriptions of archives to determine relevant topics, (2) image recognition to understand the content of an image/generate descriptions, (3) the chatbot facility, text to speech and (4) real time translation which could be useful to present a Welsh version of the service. Archival descriptions and digital content are a rich resource, and lend themselves well to Machine Learning approaches. The idea is to evaluate these experiments and work towards a more sustainable and scalable approach.