Elamur Groupantswordgraph-tamp-importance
The groupant beside wordgraph/tamp/src/importance.rs in the Mova repository: what the code computes, what it never does, how to go back from its outputs, its invariants, cost and checks.
Facts
Implementation material. A Mova code model describes this code at one source revision: partial models, finite checks and their stated scope. It is not a reviewed general exposition, and its labels are shown as written, never raised.
Components
| component | label | where |
|---|---|---|
| Approximant | partial | Approximant |
| Ast | partial | syntax.json |
| Composition | partial | Composition |
| Cost | partial | Cost |
| Descriptant | present | Descriptant |
| Invariant | partial | Invariant |
| Tests | executed bounded checks | Tests |
Obernants
The typed relations around the forward model: the behaviour complement (what the code never does), the reverse reconstruction (every input compatible with an output) and specialised views.
| relation | type | label | where |
|---|---|---|---|
| Behavior complement | behavior complement | partial source derived model | Behavior complement |
| Reverse reconstruction | reverse reconstruction | partial source derived model | Reverse reconstruction |
Model text
Descriptant
Statement importance: 0.35 rarity (statement document frequency) + 0.25 specificity (top-3 lemma idf) + 0.20 own
(self-reference with a contribution verb) + 0.15 result (number with %, metric or comparison) + 0.05 part, × 0.25
for boilerplate. The cue functions own, result, part_weight.
Approximant
score = (0.35 r + 0.25 s + 0.2 o + 0.15 n + 0.05 p)·(b ? 0.25 : 1), r = 1 − ln df / ln N (0.5 unknown).
Invariant
score ∈ [−0.05, 1]; boilerplate scales by exactly 0.25; pure function of its inputs.
Cost
O(|ids| log |ids|) per statement.
Composition
Cues used by corpus::load; score by build::scores.
Tests
own_needs_a_self_reference_and_a_contribution_verb («we use adam», «they propose» are not own — red side),
rare_specific_own_results_outrank_boilerplate.
Behavior complement
The weights are a stated operating point, not fitted; no learned model; no discourse context.
Reverse reconstruction
Each important statement in a paper record carries its rarity, specificity, own and result, so the score is recomputable.
Provenance
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/wordgraph/tamp/src/importance.rs.groupant/