Decentralised Semantics: A Semantic Engine User Perspective
Bibliographic record
Abstract
The Findable, Accessible, Interoperable and Reusable (FAIR) data principles were created to guide the improvement of research data (Wilkinson et al., 2016). As data curators and educators, we often see individual research groups and researchers establish their own unique data collection process, resulting in poor and inconsistent data documentation. At the conclusion of the project, while the data may be accessible and understood by members within the team, it is often not readily usable to anyone outside of those most closely associated with data collection and analysis. The root cause of this is the difficulty to document the pertinent information required to capture the context in which data was captured, processed, and presented. And even when this is attempted it tends to be static and non-machine actionable. As a result, the project data might be FAIR but it is not visible and the cost of re-use is too high as currently few protocols are machine actionable. The availability of context documentation will help other researchers understand and facilitate the re-use the data. Agri-Food Data Canada operates across multiple projects in different fields and run by different institutions. It is a natural environment to recognize the need of decentralized semantic definitions where each research group can influence, modify, or adjust the definition of the data while maintaining integrity of data objects (e.g., schema, data sets, catalogues) across the ecosystem. This practice paper describes the release of the first version of the Semantic Engine leveraging OCA, an architecture to document schemas optimized for decentralized collaboration and reproducibility. OCA leverages new technologies on self-addressing identifiers and enables content-based authority vs. location-based authority. We present here the first results of the Semantic Engine development and the future application.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.034 | 0.028 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.002 | 0.012 |
| Scholarly communication | 0.020 | 0.040 |
| Open science | 0.005 | 0.012 |
| Research integrity | 0.005 | 0.007 |
| Insufficient payload (model declined to judge) | 0.008 | 0.003 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".