The rubric finder tool
A rubric finder takes a symptom phrase and returns the repertory rubrics most likely to express it. It sits between the patient's spoken complaint and the analysis grid: a keyword index over a printed repertory edition, a desktop database query, or — in the cloud generation of 2026 — a vector retrieval over an embedded rubric corpus.
What a rubric finder actually does
"Rubric finder" is sometimes used as shorthand for the whole repertorisation workflow, but the finder step is narrower. Its job ends when a candidate rubric list is on the screen. Weighting, elimination, and the analysis pass are downstream tasks the practitioner — or the analysis surface — performs separately. Conflating the two is how a strong retrieval surface gets mistaken for clinical judgement.
| Approach | Input format | Disambiguation | Source coverage |
|---|---|---|---|
| Printed Kent or Boenninghausen lookup | Hierarchical chapter → section → rubric | Practitioner reads adjacent rubrics manually | One repertory edition per volume on the desk |
| Desktop keyword search (legacy software) | Exact or stemmed keyword | Boolean filters and chapter pickers | Multiple repertories per licence |
| Cloud semantic search | Natural-language symptom phrase | Nearest-neighbour ranking, with the classical hierarchy still surfaced | Complete repertory plus translated sources where licensed |
Each row trades a different cost. Printed lookup preserves the original taxonomy and asks the practitioner to do all the disambiguation. Desktop keyword search compresses lookup time but punishes phrasing variance — "fear of the dark" returns nothing if the rubric reads "darkness, fear of". Semantic search resolves the phrasing variance but introduces a ranking model whose decisions the practitioner must still audit, rubric by rubric.
How the 2026 cloud generation differs
Semantic retrieval over a rubric corpus is, technically, a well-established nearest-neighbour pattern. Its application to homeopathic rubrics is a retrieval surface — not a clinical-outcome instrument — and marketing copy that blurs that line should be read with that distinction in mind.
For a practitioner already keeping a paper repertory on the desk — or using a free online repertory for quick lookups — the practical question is whether semantic retrieval saves enough lookup time on ambiguous phrasings ("anxiety before sleep", "burning pain better from cold applications") to justify a subscription tier. Similia exposes semantic rubric search on its Free plan over the classic repertory, with translated Complete sources unlocked when the account licenses them; case-notes-to-rubrics, live audio with auto-rubric, and photo extraction sit behind the Pro tier and the AI-credit budget. Whether the recall advantage of vector retrieval holds at the long tail of rare rubrics — where the embedding training corpus is sparse — is still an open question, and one worth testing on your own corpus of difficult phrasings rather than on a vendor demo case.
Rubric retrieval is the first step in a workflow that continues through repertorisation and into remedy selection. Practitioners who want to trial the full chain alongside rubric lookup can compare results in the remedy finder, or search the free repertory directly by symptom phrase.
References
Kent, J. T. (1897) Repertory of the Homoeopathic Materia Medica, first edition.
Boenninghausen, C. M. F. von (1846) Therapeutisches Taschenbuch.
Sutton, R. T., Pincock, D., Baumgart, D. C., Sadowski, D. C., Fedorak, R. N. and Kroeker, K. I. (2020) An overview of clinical decision support systems: benefits, risks, and strategies for success, npj Digital Medicine, 3, 17.
Similia (2026) Knowledge base — Searching for Rubrics (Semantic Search), Case Notes to Rubrics, Live Audio Mode, Subscription FAQ, https://similia.crisp.help/.