An ontology and source-attributed knowledge graph for the computational heritage of Chinese mantic arts: Qimen Dunjia and Meihua Yishu
This paper presents a source-attributed ontology and knowledge graph framework that systematically digitizes and validates the contested procedural rules of Chinese mantic arts (Qimen Dunjia and Meihua Yishu) by establishing a provenance-driven verification system to preserve the auditable documentary layer of this intangible heritage.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
For centuries, a vast and intricate tradition of Chinese calculation has existed alongside the more familiar sciences of astronomy and mathematics. Known as shushu, or the "art of calculation," this system includes methods for reading the stars, tracking the seasons, and interpreting the future through complex symbolic charts. Two of its most famous branches are Qimen Dunjia and Meihua Yishu. These are not simple fortune-telling tricks but elaborate procedural systems that rely on a shared vocabulary of symbols—trigrams, stars, doors, and spirits—arranged according to strict rules. For a long time, these rules survived only in handwritten manuscripts and printed manuals passed down through generations. However, as these practices moved into the modern world, a quiet crisis emerged. Today, many people use software to generate these charts, but the code behind these programs often blends rules from different, sometimes conflicting, schools of thought into a single, seamless product. In doing so, the software erases the very differences that scholars care about, making it impossible to trace which specific version of the rules produced a given result. The practice itself remains popular, but the documentary layer that allows us to audit its history is disappearing.
A researcher has addressed this problem by building a digital archive that treats these ancient calculation methods not as mystical secrets, but as historical data that can be verified and questioned. They focused on Qimen Dunjia and Meihua Yishu, creating a structured map of the rules that govern them. Instead of trying to force all the different versions of these systems into one "correct" answer, the researcher built a system that records every disagreement. They gathered forty specific historical manuals and four modern scholarly studies, treating each edition of a book as a distinct witness. For every single rule in these systems, they checked how many independent sources supported it. If a rule appeared in at least three different manuals, it was marked as verified. If it appeared in only one, it was labeled as a single-source variation. If a rule was common in modern practice but had no support in the old texts, it was flagged as unverified. This approach turns the question of "which rule is right?" into a query about evidence: "how many witnesses back this rule?"
The result is a knowledge graph, a type of database that connects facts like a web, which now holds the computational heritage of these arts. The researcher did not just list the rules; they attached a "validation status" to each one, making the level of support for every claim a visible, searchable part of the record. To test if this system worked, they looked at a specific, long-standing debate about how to handle the center of the chart, a space that has no star or door of its own. In the past, software might have silently picked one way to fill this spot. The new graph, however, shows that six different historical manuals agree on one specific method, while a different method is supported by only a single advocate. The system records both positions side by side, clearly labeling the majority view as well-supported and the minority view as a rare variant. This allows a user to see exactly where the agreement lies and where the disagreement exists, without the software hiding the conflict.
The study also uncovered twenty-one specific points where the sources disagree on how the calculations should be performed. For sixteen of these points, the researcher found strong agreement across multiple historical texts. For the others, the evidence was thinner, or the rules were found only in modern adaptations. By documenting these twenty-one points of divergence, the researcher created a register of exactly where the tradition splits. This is crucial because it preserves the history of the practice in its true, messy form, rather than smoothing it over into a false uniformity. The work makes no claim about whether these ancient calculation methods can actually predict the future or influence real-world events. The goal was purely to preserve the rules themselves as they were written down, ensuring that the history of how these systems were constructed remains transparent and open to inspection.
This project offers a new way to handle other traditions where knowledge is passed down through different schools that disagree on specific steps. By using a system that counts witnesses and records disagreements, it protects the integrity of the source material. The researcher has made their data and the tools to read it available to the public, allowing anyone to check the evidence behind any rule. In a world where digital tools often hide the complexity of the past, this work ensures that the history of these calculation arts remains audible, traceable, and honest.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.