Relying on the annotation scheme proposed by Álvarez-Mellado et al. [1], this paper attempts to refine the proposed model for the annotation of named entities and adapt it to the needs of (medieval and early modern) epigraphy, exemplified in this article by the case of the MEMIS corpus, which brings together medieval and early modern inscriptions from the area of present-day Slovenia. Digital humanities (DH) tools and protocols provide us with ways to access and process elements of historical evidence on epigraphic monuments as documents: In addition to actual events, they include, in particular, names of persons and places. Named Entity Recognition (NER) is therefore of paramount importance for the extraction of biographical, prosopographical, etc. data. Building on the previous work of DH researchers in the field of encoding standards for humanities texts, this paper focuses on the previously unexplored medieval and early modern inscriptions in the northern Istrian (now Slovenian) towns of the former Republic of Venice.