Audit analytics researcher using hybrid rule-based and ML methods for general ledger anomaly detection
1
papers here
1
reviewed by Evolve
1
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-20) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
No publication-volume metric is available yet for this profile: output measures counts of papers, venues, and authorship roles over time, and the underlying bibliometric record currently covers too little to compute it.
Impact—
Impact cannot be scored: this axis tracks citations, downloads, and uptake of the work by other researchers, and no citation or influence data is present in the current record beyond a single 2026 preprint.
Focus100/100
Scores a maximum 100 for focus: the field distribution lists exactly one field — accounting information systems / audit analytics (applied machine learning) — with one paper. The cataloged 2026 first-author preprint sits entirely inside journal entry testing and general ledger anomaly detection. With only a single cataloged work, perfect concentration is expected rather than remarkable, and the score should be read as narrow early specialization, not established depth.
Collaboration28/100
Scores 28: the raw record lists only 3 distinct coauthors, with Alexander Michael Rombach, Sebastian Stephan, and Peter Fettke as recurring collaborators. The single cataloged paper is first-authored and non-corresponding, consistent with a small, stable working group rather than a broad network. One paper cannot show whether this tight circle is a fixed team or the early core of a wider collaboration pattern.
Activity—
No activity signal is available: this axis would measure publication cadence and recency across years, but the record holds only one 2026 item, so any estimate of ongoing research tempo would be guesswork.
Reproducibility20/100
Scores 20 (mean over 1 paper). Evolve rates the 2026 general ledger anomaly detection study low for reproducibility readiness: experiments run on synthetic audit data, and the review notes the evaluation is circular because anomalies are injected only inside JET lists. The striking false-positive reduction (10,019 to 1,881 at full recall) is therefore hard to reproduce or generalize from the published setup as described.
Research summary
Jan Gronewald is an early-stage researcher in accounting information systems and audit analytics whose cataloged work applies machine learning to journal entry testing. The Evolve catalog holds one paper: a 2026 arXiv preprint, "Anomaly Detection in General Ledger Data: Results from a Hybrid Approach," on which Gronewald is first author (non-corresponding). The study combines rule-based journal entry testing (JET) with machine learning on synthetic audit data; its headline result is an augmented isolation forest that reduces flagged false positives from 10,019 to 1,881 while preserving full recall. Evolve's review judges the result promising but methodologically circular, because synthetic anomalies are injected only inside JET lists, and rates reproducibility readiness low (reproMean 20 over one paper). The radar profile shows maximum focus (100) on a single field — accounting information systems / audit analytics (applied machine learning) — and modest collaboration (28), with 3 distinct coauthors including Alexander Michael Rombach, Sebastian Stephan, and Peter Fettke. Output, impact, and activity cannot be scored because the underlying bibliometric data is not yet available. This is a name-resolved, early-stage profile built from a thin public record.
Collaboration pattern
The raw bibliometric record shows only 3 distinct coauthors. Recurring collaborators are Alexander Michael Rombach, Sebastian Stephan, and Peter Fettke. The one cataloged paper is first-authored but not corresponding, suggesting a first-author contributor within a small, stable group rather than a corresponding-lead of a large consortium. With a collaboration score of 28 and a single data point, the pattern reads as an early-career or tightly scoped team; broader network behavior cannot yet be assessed.
Research keywords
Highlights
Augmented isolation forest cuts false positives from 10,019 to 1,881 with full recall
First-author hybrid JET-plus-ML approach for general ledger anomaly detection
Evolve review flags circular evaluation and low reproducibility readiness
Timeline
- 2026Publishes first-author arXiv preprint "Anomaly Detection in General Ledger Data: Results from a Hybrid Approach", introducing a hybrid JET-plus-machine-learning pipeline for general ledger audit data.
- 2026Reports the study's headline result: an augmented isolation forest reduces flagged false positives from 10,019 to 1,881 with full recall.
- 2026Evolve's review credits the promise but flags circularity (synthetic anomalies injected only inside JET lists) and rates reproducibility readiness low.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Dimension scores are computed by Evolve from public bibliometric data only; no citation, venue, or output aggregates are available for this profile yet.
- Focus, collaboration, and reproducibility rest on a single cataloged 2026 paper, so they describe one work rather than a track record.
- No ORCID, institution, or country is linked; the profile is name-resolved from public bibliographic data and common-name ambiguity cannot be excluded.
- Output, impact, and activity are unscored because the underlying data is missing, not because the researcher lacks publications or citations.