Schematic diagram of the ancient language processing model "Xunzi". (PHOTO: College of Information Management of Nanjing Agricultural University) |
A smart language tool for processing and research of ancient books was launched recently, making it easier to read and understand difficult ancient Chinese language.
The tool named after Xunzi, a renowned Chinese philosopher, is a language model which has collected a corpus of more than two billion words, including the "Siku Quanshu," also known as "Complete Library in the Four Branches of Literature." It can perform functions such as natural language understanding, automatic translation, poetry generation, and automatic indexing.
Thanks to its high-quality data and super computability, users can understand ancient texts without punctuation and translate the ancient expressions into modern Chinese language.
Currently, the tool is available as an open source and public welfare research result on GitHub, ModelScope and other websites.
Besides providing convenience for readers of ancient books and researchers, the "Xunzi" language model will be applied to AI writing and teaching, digital entertainment and other fields in the future, according to researchers.