Byzantine Sigillography, Linked Open Data, and the Structured Assertion Record
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
One of the major challenges in the creation of a digital dataset for cultural heritage objects is the question of how to capture the differing interpretations and the changing state of our knowledge about the objects we are seeking to document and preserve. This article discusses a case in point, namely the thousands of Byzantine lead seals that survive in museums and private collections today. The seals are an important source of information and insight into a society whose copious administrative records we have by and large lost; their correct decipherment and interpretation, however, requires a particular expertise that few in the field possess, and the need for a more or less central source of information about these seals has been acknowledged for many years now. As part of the Prosopography of the Byzantine World (PBW) project, a database was created that aimed for as complete a coverage as possible of all seals dated to the eleventh and twelfth centuries, including an innovative organization of the seals according to the boulloterion (die) from which a particular seal was struck and a link between the boulloterion and its owner. The strength of this database is that it is a rich collection of sigillographic data unparalleled elsewhere; the weakness is one shared with almost every digital database in the historical sciences, specifically, that it presents a single interpretation of the data when multiple interpretations are possible.The aim of the RELEVEN project has been to re-think how databases of historical information are structured; its central innovation is the “structured assertion record” (STAR) model, which is a Linked Open Data model based on the CIDOC-CRM standard. Here we discuss how the STAR model has been applied to the PBW seals database to express the information in a CIDOC-CRM-conformant way, and also to preserve information in all cases about who has made a particular interpretation of the data and what source material was used for the interpretation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it