The research is about the offline handwritten character recognition system using the BERT model. The collective data that is being obtained here can be very helpful in the case of implementing different tools and processes in this model. The BERT models framework has already been preinstalled using different text from Wikipedia and also has been fine tunes by the inclusion of different datasets that have questions and answers. Different types of problems were found in the implementation of the newly developed natural language model that was being developed by BERT as well as the problems of using an offline handwriting character recognition system. The solution to this problem has also been provided using different artificial intelligence-based machine learning and deep learning algorithms.

BERT model, Offline handwritten Character, Algorithm