Third Workshop on Ancient Language Processing
Welcome to ALP 2027!
Ancient languages preserve humanity’s earliest recorded intellectual and cultural achievements, yet they remain among the most challenging targets for language technologies. While significant progress has been made in developing digital resources and computational methods for ancient languages, many challenges remain, including scarce and fragmentary data, damaged textual sources, limited expert annotation, and complex writing systems that are often non-Latin, non-alphabetic, or lack stable encoding conventions.
The Third Workshop on Ancient Language Processing (ALP 2027) focuses on ancient languages and scripts from the emergence of writing circa 3400 BCE to cultures worldwide up to 800 CE. Building on ALP 2023 and ALP 2025, the workshop provides a forum for researchers and practitioners to exchange ideas and advance NLP for ancient languages. ALP 2027 welcomes research on evaluation under data scarcity, multimodal and visual script processing, language models for philological workflows, data governance and reproducibility, as well as related topics including OCR, transliteration, corpus construction, linguistic analysis, machine translation, and the processing of uncertain and fragmentary texts. The workshop also features two shared tasks: EvaCun 2027, focusing on end-to-end OCR for cuneiform languages, and EvaHan 2027, addressing ancient Chinese text difficulty grading.
Languages of interest include, but are not limited to:
- Mesopotamia: Sumerian, Akkadian
- Iran: Elamite, Old and Middle Persian
- Levant: Eblaite, Amorite, Aramaic (incl. Mandaic and Syriac), Ancient Hebrew, Phoenician, Ugaritic
- Anatolia: Hittite, Luwian and minor Anatolian languages
- Egypt: Ancient Egyptian, Coptic
- Mediterranean: Linear A and B, Ancient Greek, Latin
- Arabia: Ancient North Arabian, Old Arabic
- South Asia: Sanskrit, Eastern Punjabi, Pali
- East Asia: Literary Chinese, Tibetan, Old Japanese
- Mesoamerica: Mayan