Multiple Views of Knowledge in Diagnosis
Bibliographic record
Abstract
This paper argues that automated knowledge acquisition for diagnosis has had limited success in both failure-driven diagnosis and model-based diagnosis. The paper describes fault-based and model-based reasoning for diagnosis and surveys some of the approaches to knowledge acquisition in both areas. The Diagnostic Remodeller (DR) algorithm I am currently implementing for the automated generation of behavioural models from fault-based knowledge is presented. An example of fault-based knowledge from the Jet Engine Troubleshooting Assistant (JETA) is used to demonstrate how a behavioural model can be extracted with DR. Fault-Based Diagnosis Fault-based reasoning (FBR) is used in many diagnostic systems. Knowledge in FBR is largely based on maintenance manuals and interviews with experts intended to capture heuristic knowledge about the maintenance and repair of a device or process. The knowledge in these systems is often represented as handcoded rules or frames which are organized into troubleshooting hierarchies. At the top level of the hierarchy is the general knowledge representing a problem with the device. This general problem is refined systematically until the leaf nodes of the hierarchy which represent physical repairs to the device are reached. Once these repairs are achieved by a human technician some diagnostic systems re-test to confirm that the symptoms and diagnosed faults are cleared through backtracking in the hierarchy. FBR systems have evolved considerably since the development of MYCIN [Scott et al. 77]. MYCIN was developed to provide advice treatment for microbial infections. The MYCIN programs started with handcoded rules which later evolved into meta-rules in NEO-MYCIN to provide some structure to an otherwise flat
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 | 0.018 | 0.020 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.008 | 0.004 |
| Science and technology studies | 0.004 | 0.036 |
| Scholarly communication | 0.019 | 0.040 |
| Open science | 0.004 | 0.013 |
| Research integrity | 0.007 | 0.009 |
| Insufficient payload (model declined to judge) | 0.008 | 0.001 |
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".