Word-level Alignment and Named Entities in the Trilingual Inscription at Ka'ba-ye Zartošt (ŠKZ)

This dataset includes the corpus of Greek, Middle Persian, and Parthian versions of the inscription aligned at both sentence and word levels, and manually extracted named entities. The corpus follows the line numbering of Huyse (1999). The Greek text is taken from the digital epigraphy collection of the Packard Humanities Institute, which uses the edition of Canali De Rossi (2004).The Parthian and Middle Persian versions are based on Huyse's edition. The Middle Persian version was digitized for the purpose of this project, and the Parthian version is taken from Jake Nabel's digital resource at http://parthiansources.com.The alignments were produced using the Ugarit alignment tool. All alignments are available openly online on Ugarit here: https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=21881The extracted alignment pairs are made available both as one file (xslx, csv and tsv) and also as seperate files (csv and xlsx) in zip, where the name of each file is the line number.

The named entity dataset includes nearly 400 named entities that were extracted and classified manually as persons (PER), locations (LOC), or location derivatives (LOCderiv). The dataset contains the named entities across all three versions (if available) and the line number in which they appear. A few also include a reference to Wikidata.

Sources:

Canali De Rossi (2004): F. Canali De Rossi (ed.), Iscrizioni dello estremo oriente greco. Un repertorio. «Inschriften griechischer Städte aus Kleinasien» 65, Bonn 2004.Huyse (1999): P. Huyse (ed.), Die dreisprachige Inschrift Šābuhrs I. an der Kaʿba-i Zardušt (ŠKZ), Corpus Inscriptionum Iranicarum III/I/I, London 1999.