Mahdi Hajiali (Computer Science) will have a new study appear in the Proceedings of the 2021 Computing Conference, a highly ranked conference in the areas of computer science and computing. Collaborators included Mahdi Hajiali, Jorge Ramón Fonseca Cacho, and Kazem Taghva (all Computer Science). The study, “Generating Correction Candidates for OCR Errors using BERT Language Model and FastText SubWord Embeddings” utilizes the neural language models as a part of Google's BERT natural language processing tools. Mahdi's work, which deals with applications of machine learning in detections and corrections of errors caused by optical character recognition, is heavily dependent on computing resources that were obtained from a National Science Foundation grant.