Dataset
big_patent
Over 1.3 million granted US patents from the BIGPATENT benchmark, each pairing its claims with an abstractive summary and full description text in English (CC BY 4.0, Parquet). Use it to track the technology areas where applicants are filing and to study how claim language has evolved over time.
Data preview
Dataset card
Big Patent is a large public corpus of U.S. patent documents released by Northeastern University, widely used to train abstractive summarization models. It pairs long patent descriptions with human-written abstracts across nine Cooperative Patent Classification (CPC) categories.
What this dataset contains
The dataset holds approximately 2.68 million rows of U.S. patent texts, each pairing a long-form technical description with a shorter abstract. Patents are grouped under nine Cooperative Patent Classification (CPC) codes, which users can load individually (for example, the 'a' subset) or combined as 'all'. Content is drawn from original U.S. patent filings and is distributed in Parquet format.
Structure and fields
Each row exposes two string fields: a `description` field containing the text from the patent's Description section, and an `abstract` field containing the corresponding human-written abstract. The corpus is partitioned into standard train, validation, and test splits. A v2.1.2 release provides cased raw strings, while the v1.0.0 revision offers lower-cased, tokenized text accessible via a pinned dataset revision hash.
How analysts use it
Analysts typically load specific CPC subsets such as `big_patent/a` to study technology areas of interest and to fine-tune summarization models that condense long patent descriptions into short abstracts. The pairing of `description` and `abstract` makes it useful for tracking the language of new filings, prototyping prior-art search, or building text analytics over public patent records. Because the underlying records are public filings, the dataset can also support IP trend monitoring as a component of broader alternative-data pipelines.
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.