[SoK] Evaluations in Industrial Intrusion Detection Research
Bibliographic record
Abstract
Industrial systems are increasingly threatened by cyberattackswith potentially disastrous consequences. To counter suchattacks, industrial intrusion detection systems strive to timelyuncover even the most sophisticated breaches. Due to its criticality for society, this fast-growing field attracts researchersfrom diverse backgrounds, resulting in 130 new detectionapproaches in 2021 alone. This huge momentum facilitatesthe exploration of diverse promising paths but likewise risksfragmenting the research landscape and burying promisingprogress. Consequently, it needs sound and comprehensibleevaluations to mitigate this risk and catalyze efforts into sustainable scientific progress with real-world applicability. Inthis paper, we therefore systematically analyze the evaluationmethodologies of this field to understand the current stateof industrial intrusion detection research. Our analysis of609 publications shows that the rapid growth of this researchfield has positive and negative consequences. While we observe an increased use of public datasets, publications stillonly evaluate 1.3 datasets on average, and frequently usedbenchmarking metrics are ambiguous. At the same time, theadoption of newly developed benchmarking metrics sees littleadvancement. Finally, our systematic analysis enables us toprovide actionable recommendations for all actors involvedand thus bring the entire research field forward.
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 | 0.033 | 0.146 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.019 | 0.024 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.006 | 0.007 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.010 | 0.005 |
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".