I. Introduction
Historical records of daily activities provide intriguing insights into the life of our ancestors, useful for demography studies and genealogical research [1]. Handwritten marriage licenses books [2] are one of these records that have been used for centuries by ecclesiastical and secular institutions to register marriages. The information contained in these historical documents is very interesting for migratory studies, population research and genealogical investigation. Therefore, one of the goals of this kind of documents, rather than to transcribe perfectly the documents, is to extract the relevant information to allow the users to make use of it through semantic searches. Note that, if the perfect transcript is obtained, then identifying the relevant semantic information would be much easier, but it is not mandatory to obtain the perfect transcript.