Dataset
uspto-patent-data
A collection of Parquet files providing one record per Chinese invention patent family granted between 2003 and 2019, capturing a cumulative quality assessment pipeline that includes metrics such as semantic knowledge recombination, strict historical semantic novelty, and IPC-based robustness measures across global patent publications.
Data preview
Dataset card
The uspto-patent-data dataset provides a normalized, Parquet-formatted mirror of roughly 2.7 million U.S. patent application records published by the USPTO Patent File Wrapper between 2021 and 2026. It is hosted on the trentmkelly repository as a Hugging Face dataset and is intended for large-scale text analysis of public patent prosecution data.
What this dataset contains
The dataset mirrors public U.S. patent application records released through the USPTO Patent File Wrapper. The single training split holds 2,707,938 rows covering annual source files from 2021 through 2026. Records combine application metadata, applicant and inventor information, classifications, prosecution events, continuity and priority data, assignments, publications, and grants where available, with optional fields preserved as null values.
Structure and fields
Records are distributed as Zstandard-compressed Parquet files derived from the original USPTO JSON. Top-level columns include applicationMetaData, applicationNumberText, assignmentBag, childContinuityBag, correspondenceAddressBag, eventDataBag, foreignPriorityBag, grantDocumentMetaData, lastIngestionDateTime, parentContinuityBag, patentTermAdjustmentData, pgpubDocumentMetaData, recordAttorney, source_file, and source_year. Nested structures such as applicantBag, assigneeBag, and eventCode are preserved within the corresponding parent columns.
How analysts use it
Because it is sourced from a U.S. public-records portal, the data is suited to identifying patenting activity, applicant and assignee ownership changes, and prosecution event histories. Researchers can group records by source_year or source_file to study filing trends, use applicationNumberText and parentContinuityBag to trace family relationships, and parse eventDataBag to model prosecution timelines. The license is not declared, so commercial use should be confirmed against USPTO Open Data Portal terms before integration.
Documentation & references 1
Related datasets
Browse moreParsed eligibility criteria from 13,229 ClinicalTrials.gov studies representing the candidate pool gathered by two first-stage retrievers during TREC Clinical Trials 2021–2023, formatted as typed entity-relation graphs for reranking models.
The clinical-trials-patient-graphs dataset is a small tabular collection of patient-level clinical records distributed in Parquet format by the 2001jdev publisher. It organizes entities and relations extracted from trial data into structured rows spanning 2021 through 2023.
This dataset is a parsed snapshot of ClinicalTrials.gov records, distributed in Parquet format on the Hugging Face Hub under an unspecified license. It contains roughly 52,000 trial entries drawn from public registry filings.
clinical-trials-trec-qrels is a tabular relevance-judgment file distributed by the 2001jdev user on the Hugging Face Hub. It maps clinical-trial topic identifiers to NCT registry IDs with graded relevance scores.
The clinical-trials-trec-topics dataset is a small tabular collection of clinical-trial topic records distributed in Parquet format on the Hugging Face Hub. It appears to be derived from TREC clinical-trial retrieval topics, with rows partitioned by year.
SEC EDGAR 10-K & 10-Q financial statements for 2,600+ NASDAQ companies.
Discussion
Questions or observations about this dataset? Join the conversation on X.
No comments yet. Start the conversation on X.