I maintain this Actor. This is an AI-assisted technical walkthrough, fact-checked against its accepted source contract and saved acceptance evidence.
Citation exports often contain both observed arrays and source-reported totals. An array with three entries and a reported total of twenty is a partial view. Treating it as the complete citation set can distort a comparison just as much as treating a missing array as an observed zero.
The Patent Citation Overlap Analyzer compares supplied patent publication records. It keeps backward and forward citation directions separate, deduplicates repeated observations, and reports bounded peer overlap evidence. It does not fetch a fresh citation graph or evaluate prior art, patent validity, infringement, or patent quality.
Keep publication identity exact
Patent and citation identifiers use normalized publication identifiers with a country prefix and kind code. Kind codes remain identity-significant. Publications from the same family are not silently merged. If an application needs family-level analysis, that is a separate transformation requiring explicit family evidence and its own counting rules.
Repeated source rows should not inflate a publication's citation set. Contradictory observations remain explicit instead of being overwritten. This matters when combining source exports captured at different times: a smaller array may represent a partial capture, rather than evidence that citations disappeared from the underlying network.
Two directions deserve two measurements
Backward citations and forward citations represent different directed relationships. The Actor does not collapse them into one vague similarity score. Inspect backwardCitationCount, forwardCitationCount, the source-reported totals, and their corresponding partial-coverage flags before interpreting peer overlap.
A valid empty array represents an observed zero within that source observation. An absent or malformed array is unknown and cannot become zero. Self-citation observations are also explicit. Keeping these distinctions in a downstream table makes later filtering reproducible instead of relying on undocumented cleaning rules.
Read a no-overlap action narrowly
The saved acceptance example includes citation_overlap_assessed, action no_overlap_in_batch, and high confidence. That action refers to the supplied batch and its supported comparisons. It does not establish that the publication has no related patents, no shared prior art, or no overlap in a complete citation graph.
Useful peer fields include topOverlappingPeers, overlappingPeerCount, overlappingPeerCountIsLowerBound, and peerEvidenceOmitted. Pair comparisons are bounded to at most one thousand patents, while evidence display has its own bounds. A lower-bound flag matters when turning the result into a chart or ranking; it should survive export next to the count.
For acceptance, use a duplicated publication, two peers sharing a known citation, and a row whose reported total exceeds its observed array length. Add separate empty and missing-array cases in clearly labelled synthetic fixtures. Confirm that duplicate rows do not increase the metric, direction remains visible, and partial coverage stays explicit. Inspect record conflicts before comparing results across snapshots. The output is a batch review aid with evidence, not a measure of legal strength or a complete map of the patent landscape.
Run a small saved snapshot
Start with an existing Dataset you are authorized to read. Replace the Dataset placeholder below with its actual ID. Supported input can also use a source run ID or inline records; choose one explicit source. This Actor does not launch the upstream collection for you.
curl --request POST \
'https://api.apify.com/v2/acts/zinin~patent-citation-overlap-analyzer/runs?maxTotalChargeUsd=0.10' \
--header "Authorization: Bearer $APIFY_TOKEN" \
--header 'Content-Type: application/json' \
--data '{"datasetId":"REPLACE_WITH_YOUR_DATASET_ID","maxInputRows":100,"maxResults":10}'
The response identifies an asynchronous run. Save that run ID and check its status before reading its default Dataset. A client timeout does not cancel a server run; inspect the existing ID before starting another one. maxInputRows bounds ingestion and maxResults caps useful output. Neither option establishes complete coverage of a larger source. The API charge limit is a ceiling, not a promise that ten results will fit.
Price and evidence limits
On 15 October 2026 the observed FREE-plan price is $0.04 per citation assessment, with no paid start event for this Actor. Other subscription tiers can have different prices; verify the current Actor page before use. Source collection is a separate activity and may carry its own charges. Free diagnostic rows do not count as paid useful assessments.
The examples discussed here come from saved owner acceptance runs. They show specific observed behavior, rather than customer acceptance, profitability, or an exhaustive guarantee for every input. A new run of identical input can be charged again without adding new source evidence. Retain input provenance, the actual build ID, and the source observation window when comparing results over time.
Top comments (0)