Authors: Farhan Farsi, Parnian Fazel, Sepand Haghighi, Sadra Sabouri, Farzaneh Goshtasb, Nadia Hajipour, Ehsaneddin Asgari, Hossein Sameti Venue: Ancient Language Processing Workshop (ALP) at NAACL, 2025
Summary
Historical languages are hard to work with, and for several reasons at once: the orthography is complicated, what survived of the written record survived only in fragments, and there is no standard way to represent any of it digitally. Standard NLP tooling assumes none of those problems.
ParsiPy is a Python toolkit for reading historical Persian in spite of them. It provides modules for tokenization, lemmatization, part-of-speech tagging, phoneme-to-transliteration conversion, and word embeddings.
Links
Related
PahGen: Generating Ancient Pahlavi Text via Grammar-Guided Zero-Shot Translation — same team, adjacent problem in Pahlavi. naab: A Ready-to-Use Plug-and-Play Corpus for Farsi — the modern-Persian counterpart in the same line of Persian-language resource building.