Pythia is an ancient text restoration model that recovers missing characters from a damaged text input using deep neural networks. It was created by Yannis Assael, Thea Sommerschield, and Jonathan Prag, researchers from

Google DeepMind DeepMind Technologies is a British artificial intelligence subsidiary of Alphabet Inc. and research laboratory founded in 2010. DeepMind was acquired by Google in 2014 and became a wholly owned subsidiary of Alphabet Inc, after Google's restru ...

and the

University of Oxford , mottoeng = The Lord is my light , established = , endowment = £6.1 billion (including colleges) (2019) , budget = £2.145 billion (2019–20) , chancellor ...

. To study the society and the history of ancient civilisations,

ancient history Ancient history is a time period from the beginning of writing and recorded human history to as far as late antiquity. The span of recorded history is roughly 5,000 years, beginning with the Sumerian cuneiform script. Ancient history cove ...

relies on disciplines such as

Epigraphy Epigraphy () is the study of inscriptions, or epigraphs, as writing; it is the science of identifying graphemes, clarifying their meanings, classifying their uses according to dates and cultural contexts, and drawing conclusions about the wr ...

, the study of ancient inscribed texts. Hundreds of thousands of these texts, known as inscriptions, have survived to our day, but are often damaged over the centuries. Illegible parts of the text must then be restored by specialists, called

epigraphist Epigraphy () is the study of inscriptions, or epigraphs, as writing; it is the science of identifying graphemes, clarifying their meanings, classifying their uses according to dates and cultural contexts, and drawing conclusions about the wr ...

s, in order to extract meaningful information from the text and use it to expand our knowledge of the context in which the text was written. Pythia takes as input the damaged text, and is trained to return hypothesised restorations of ancient Greek inscriptions, working as an assistive aid for ancient historians. Its

neural network A neural network is a network or circuit of biological neurons, or, in a modern sense, an artificial neural network, composed of artificial neurons or nodes. Thus, a neural network is either a biological neural network, made up of biological ...

architecture works at both the character- and word-level, thereby effectively handling long-term context information, and dealing efficiently with incomplete word representations. Pythia is applicable to any discipline dealing with ancient texts (

philology Philology () is the study of language in oral and writing, written historical sources; it is the intersection of textual criticism, literary criticism, history, and linguistics (with especially strong ties to etymology). Philology is also defin ...

papyrology Papyrology is the study of manuscripts of ancient literature, correspondence, legal archives, etc., preserved on portable media from antiquity, the most common form of which is papyrus, the principal writing material in the ancient civilizations ...

codicology Codicology (; from French ''codicologie;'' from Latin , genitive , "notebook, book" and Greek , '' -logia'') is the study of codices or manuscript books. It is often referred to as "the archaeology of the book," a term coined by François Masai. ...

) and can work in any language (ancient or modern).

References

{{reflist Machine learning Digital humanities projects Digital humanities Epigraphy Ancient history