Online research platform Oracle Bones Corpus welcomes contributions from researchers and enthusiasts

Tencent today announced the launch of Oracle Bones Corpus (殷契文渊), an online research platform and database for oracle bone researchers and enthusiasts. This revamped website not only offers new innovative tools for visualizing shallow indentations, but also employs artificial intelligence (AI) to enhance the efficiency of mapping glyphs to modern Chinese characters. Available in Simplified Chinese, the platform is now live at https://www.jgwlbq.org.cn/home.

First discovered in the early 20th century, oracle bones record some of the oldest known iterations of Chinese characters, dating back millennia. To date, around 160,000 pieces of oracle bones have been discovered, from which around 4,500 unique characters have been found. While significant progress has been made, deciphering the scripts remains a challenge. Among all the characters which have been found, less than half (approximately 1,500 characters) have been matched to their modern counterparts.

Oracle Bones Corpus aims to a provide mature, standardized system and technological tool for conducting effective research. Its database lets researchers examine and compare – either side-by-side or layered on top of each other in an interactive viewer – the different digitized versions of oracle bones, including historical photos, high-resolution 3D models, as well as digital replicas and enhanced ink rubbings.

With the new update, Oracle Bones Corpus introduces tools that directly address the main difficulties associated with studying oracle bones: visualization and character recognition. Because of their fragility, the priceless oracle bones are often prohibited from being handled. The pieces are also scattered across China and the world, further restricting access. For a long time, researchers have had to depend on Chinese ink rubbings and photographic copies, but the quality of the records produced by these methods falls short of modern expectations. Tencent’s Indentation Highlight feature offers a solution by enhancing the visibility of shallow carvings on the oracle bones. This allows for a detailed analysis of the strokes that form the glyphs and the chiseled grooves at the back.

For character recognition, researchers previously had to laboriously compare each image and sometimes even resort to hand-drawn notes to find recurring glyphs. This AI assistance significantly boosts the efficiency and accuracy of deciphering ancient Chinese text. Additionally, characters are indexed for easy searchability, allowing researchers and interested users to quickly locate relevant academic papers for further study. By harnessing the technology, researchers can greatly shorten the processing time of oracle bones.

Commenting on the initiative, Zhan Shu, Head of Digital Culture Lab, Tencent's Sustainable Social Value Organization, said, “With Oracle Bones Corpus, we have established a platform for all the stakeholders in oracle bone research to work together toward a common goal. We look forward to seeing more oracle bones digitized and invite researchers and anthropology enthusiasts to decipher the secrets of these ancient texts.”

In promotion of oracle bone research, Tencent also co-developed the Amazing Oracle Bones (了不起的甲骨文) educational Weixin Mini Program. Through fun, interactive experiences such as word of the day, 3D model viewing and oracle bone trivia quizzes, the Mini Program aims to raise public awareness and interest in oracle bone research. Additionally, users may also submit their interpretations of unidentified characters to researchers through the Mini Program and contribute to the on-going deciphering efforts.

Oracle Bones Corpus (https://www.jgwlbq.org.cn/home) and Amazing Oracle Bones are both live and open to the public. To access the Amazing Oracle Bones Mini Program, please scan the following code with the Weixin app.