Extracting text from Cairo Genizah manuscripts

The Cairo Genizah manuscripts sketch a 1,000-year continuum of
Middle-Eastern history. Digitization of these manuscripts is making
them accessible to researchers in a wide variety of disciplines
worldwide. We survey some of the challenges that arise and show how
natural language methods are used to help in tasks such as language
identification, language modeling and approximate string matching.