Cham Documentation – CHAMDOC

Cham Documentation
Keywords : Vietnam, Cham, ancient cham, middle cham, automatic indexing of the cham characters, artificial intelligence.
Parterss : CASE, Paris ; L3i, La Rochelle ; MICA, Hanoi ; Lab. des Structures Sonores, Ottawa.
Fundings requested : 418219 €
Duration : 48 months

The CHAMDOC project focuses on one observation: in Vietnam, due to a lack of transmission, the written Cham heritage is on the verge of disappearing, whether it concerns inscriptions or the oldest manuscripts. Because of the cultural rupture of the second half of the 20th century in Vietnam, the country did not train young scientists in epigraphy and Cham palaeography and thus, the new generation of researchers who take up historical studies is cut off from this type of sources. Since 1832, the Chams have been being assimilated by the Vietnamese state, whose writing has been exclusively Romanized since 1920, and alpha-syllabic languages such as Cham are disappearing and, with them, a whole panoply of the country's history and culture. However, the Cham people have had a rich and eventful history between their blossoming in kingdoms on the coasts of Vietnam from the beginning of the 1st millennium to 1471, then their contraction in a small kingdom in Southern Vietnam until 1832, and finally their dispersion in the diaspora. Each stage of its history has been accompanied by texts reflecting its socio-political-religious functioning: inscriptions in Sanskrit and Ancient Cham from the 6th to the 15th century; manuscripts in Middle Cham on latanier leaves or on paper from the 18th century; manuscripts in Modern Cham during the French colonial period. The project aims to provide Vietnam's history with the opportunity to access these documents in the languages that have now disappeared: the Ancient and the Middle Cham.
Thanks to the close collaboration between researchers in SHS and IT, it is now possible to reverse this fatality and propose a project on the automatic recognition and analysis of Cham characters and Cham words in two types of documents: inscriptions and manuscripts. In order to avoid the disappearance of these alpha syllabic writings, the CHAMDOC project aims to develop methods of image processing, analysis and recognition, based on artificial intelligence techniques, to allow automatic reading of sources (inscriptions in Ancient Cham and Sanskrit until the 17th century, Middle Cham manuscripts from the Royal Archives of Panduranga between 1702 and 1850), then their indexation for wider use.
The CHAMDOC project is a multidisciplinary, unique and innovative project. Its overall objective is the design and development of innovative digital methods and tools for extracting and analysing characters from old written documents. More specifically, four objectives will be pursued:
- the collection, analysis and indexing of documents written in Ancient and Middle Cham;
- the development of image processing and pattern recognition methods based on artificial intelligence techniques to analyse the structure and content of Cham inscription and manuscript images;
- indexing of content to allow the creation and use of semantic links between documentary sources;
- the permanent safeguarding and enhancement of a unique cultural heritage.
The project partners are computer research groups specialized in automatic content extraction (L3i and MICA) and SHS research on Cham (CASE and LSS). Together, they want to bring the written Cham heritage into the digital age and give a new impetus to this documentation. The project is positioned in the field of digital humanities, combining SHS and Computer Sciences. It is part of the scientific axis "Culture, creations, heritage". It aims to preserve the history and cultural heritage of the Cham people in Vietnam.