Alternative data datasets
Browse by provider, category, marketplace, access, delivery and license
View source information and access options.
Complimentary 2019 preview offering high-level foot-traffic patterns across sectors and geographies over time.
View source information and access options.
View source information and access options.
View source information and access options.
Parsed 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.
The job-postings-english-clean dataset, published by 2024-mcm-everitt-ryan, is an English-language corpus of online job postings distributed as a single Parquet train split of roughly 1.76 million rows. It is tagged as US-region data and is intended for text-based machine learning workflows.
The job-postings-raw dataset is a large-scale tabular collection of scraped job postings distributed in Parquet format. It aggregates records across multiple sources and locales, with US coverage indicated in the dataset metadata.
Humanoid On-Chain Interaction Events This dataset captures interactions between humanoid agents and blockchain systems. Designed to bridge humanoid intelligence with on-chain logic and decentralized coordination. Contents Transaction intent Smart contract interaction On-chain result Use Cases Autonomous agents On-chain AI behavior Web3 coordination Part of Humanoid Network (HAN) License MIT
Datamule, Teraflop AI, and Eventual jointly contributed a 590 gigabyte SEC EDGAR collection encompassing about 8 million filings and 43 billion tokens, assembled using the datamule-python library and the official datamule API.
A sharded Parquet archive of NSE equities and index prices from India spanning 2000 to 2026, covering more than 2,500 tickers and organized into roughly 1.5 GB files for efficient streaming.
credit_card_transactions is a small tabular dataset published by aegisheld containing anonymized customer-level credit card account and spending summaries. It is distributed as CSV with a single training split of 8,950 rows.
About 1.5 million US patent claims split into training and test partitions in CSV, organized for claim-level classification and summarization work (Apache 2.0). A large, ready-split corpus for IP-claims modeling.
A component of the ALEA Institute's KL3M Data Project supplying training material cleared of copyright concerns. The dataset card is currently a placeholder, directing readers to the GitHub repo and project paper for full documentation.
The KL3M Data Project from the ALEA Institute supplies copyright-clean training material, and the dataset listed here is one component pending further documentation on its dedicated page.
Material contracts and agreements extracted from EDGAR filings by the KL3M project: debt, M&A, employment and license agreements, millions of documents in Parquet. Contract language is a niche but real input for M&A, financing and litigation-event signals.
The kl3m-index-edgar-filings dataset, published by ALEA Institute, is a Parquet-format index of SEC EDGAR filings distributed via the Hub datasets library. It contains roughly 20 million tabular records spanning the 10M–100M size category, with an unstated license.
The KL3M Index of EDGAR 10-K Filings is a tabular index published by the Alea Institute that catalogs SEC EDGAR filings with associated issuer metadata. It is distributed in Parquet format and is tagged for use with pandas, polars, and the mlcroissant dataset library.
The kl3m-index-edgar-filings-8-k dataset is an indexed catalog of Form 8-K filings from the SEC's EDGAR system, published by the ALEA Institute. It provides structured metadata for over 1.8 million corporate event reports distributed in Parquet format.
US patent full text assembled by the Allen Institute from the USPTO, millions of records under ODC-BY: claims, abstracts, descriptions and metadata, in Parquet. A broad base for measuring IP intensity, technology landscapes and the direction of R&D spending.
A private collection of cleaned earnings-call transcript segments derived from a Motley Fool Kaggle release, split into 135,306 training rows tagged with ticker, exchange, date, and chunk text.
A static tabular extension of the Bose345/sp500_earnings_transcripts collection covering the same 2005 to 2025 calendar of S&P 500 earnings events, where each row represents a single company-quarter call identified by a stable episode_id and bundles the full transcript along with related SEC press materials and pre-event inputs suitable for supervised learning or reinforcement-style experimentation.
EDGAR_FILINGS_DATASET_2016_2021 is a Hugging Face mirror of parsed SEC EDGAR filings spanning 2016 through 2021, distributed as a single train split of about 6 million rows in Parquet format. It is published by anonymous-md with an unstated license.
EDGAR_FILINGS_DATASET_2022_2026H1 is a Hugging Face dataset that compiles parsed SEC EDGAR filings into a single tabular corpus. It covers roughly 2022 through the first half of 2026 and is distributed as Parquet with about 1.7 million records.
The TREC Clinical Trials collections released in 2021, 2022, and 2023 are available through the TREC homepage, with corresponding papers for each year. Artur Guimarães ([email protected]) serves as the dataset curator, while inquiries regarding the original authors can be directed separately. The dataset addresses the task of linking patient profiles with appropriate clinical trials, containing entries formatted as JSON objects with fields including query identifiers, disease labels, and accompanying clinical text.
An aggregation of 1.3 million U.S. patent records each accompanied by a human-written abstractive summary, sorted into nine Cooperative Patent Classification groupings spanning human necessities, chemistry, textiles, and related domains.
The awinml/earnings_calls_transcripts dataset on the Hugging Face Hub packages a small collection of earnings-call transcript segments formatted as chat-style messages for fine-tuning language models. It is distributed as a parquet file and is tagged for text modality with libraries including datasets, pandas, mlcroissant, and polars.
Identify high-spending consumers across 10 categories
The airline-otp-data dataset is a large-scale, public-style tabular repository of U.S. domestic flight records covering on-time performance, delay attribution and cancellation outcomes. Published on Hugging Face under an unstated license, it contains roughly 30 million rows in CSV format and is aimed at analysts working with airline operations data.
A large-scale release produced jointly by Datamule, Teraflop AI, and Eventual covering roughly 590 GB of SEC EDGAR content. It comprises about eight million filings totaling 43 billion tokens, gathered using the datamule-python package and the official datamule API.
This dataset is a filtered version of https://huggingface.co/datasets/vincha77/filtered_yelp_restaurant_reviews
S&P 500 earnings-call transcripts in Parquet under MIT, tens of thousands of calls. Independently compiled and comparable to other public transcript corpora for sentiment and tone work.
This dataset has moved to anamnesis-data/clinical_trials_history. New releases are published there only; this repository is no longer updated. Access approvals do not carry over, so please request access on the new repository. Existing tags here (up to v2026.09.26) remain available for pinned use. ClinicalTrials.gov Version History A longitudinal, per-version snapshot of every protocol amendment e
A reference collection drawn from the U.S. Patent Phrase to Phrase Matching Kaggle competition, containing additional details that are still required for full documentation.
A sample of US patent titles, abstracts and CPC classification labels built for multi-class patent classification, tens of thousands of text records in Parquet. Useful as training material for mapping innovation activity onto technology categories over time.
clinical-trials-v2 is a Hugging Face dataset published by chemNLP containing processed ClinicalTrials.gov records stored in Parquet. It exposes three columns, filename, xml, and text, drawn from the official clinical-study XML schema.
Chinese patents paired with their abstractive summaries in Mandarin, a few thousand records in JSON (Apache 2.0). A window into the pace and direction of patenting inside the Chinese technology base.
A collection of earnings call transcripts covering S&P 500 firms and other U.S. large-cap companies across the years 2005 through 2025, intended for financial analysis, NLP modeling, and sentiment research.
A small text-classification dataset by ClarusC64 that labels whether borrow-rate, short-interest and price signals cohere into a real short squeeze or diverge into a false signal. It is published on a model hub with an MIT license tag and a 10-row train split.
GenAI-job-postings-Dataset is a small, US-focused text corpus of generative-AI and machine-learning job postings distributed in Parquet format. The dataset is published on the Hugging Face Hub under an unstated license and comprises a single train split of roughly 120 rows.
A collection of restaurant reviews was assembled in 2019 through Python-based web scraping focused on Dutch establishments, capturing both visit experiences and feature-related information. It is organized in the DatasetDict format with three splits: 116,693 training records, 14,587 test records, and 14,587 validation records.
Mechanical Utility Patents — 100 HTML Records A curated collection of 100 distinct granted U.S. utility patents from 100 patent families, saved as HTML for mechanical-engineering document analysis. Published by cmuchancel. Source disclosure: the patent text, bibliographic metadata, and drawings were obtained from Google Patents. Original PDFs and drawing assets are hosted on patentimages.storage.g
OnchainPlanBench Seed OnchainPlanBench Seed is an early CodePit dataset for evaluating whether small open-weight models can critique, repair, reject, or approve Web3 AI-agent action plans before wallet execution. This is not a production safety benchmark yet. It is a public seed release for the first CodePit official model track: CodePit PlanGuard. What It Tests Each row gives a model: user intent
Records of initial public offerings launched on Indian markets between 2006 and 2025, with fields covering open and close dates, listing date, face value, issue price and size, lot size, first-day price, total shares offered and their allocation across anchor, NII, QIB and retail categories, minimum investment, and subscription figures for each investor class.
sec-edgar-rpt is a Hugging Face dataset published by codewithdark that aggregates SEC EDGAR filings, packaged as a single text column with a train split of roughly 3.9 million rows. It is distributed under an unstated license and is catalogued for text and machine-learning workflows.
A tabular extract of clinical study descriptions drawn from ClinicalTrials.gov on 5 February 2025, emphasizing eligibility, design, and objective fields for healthcare and machine learning research.
Vector embeddings of clinical-trial records, hundreds of thousands of entries in Parquet (Apache 2.0). Enables similarity and retrieval analysis of trial design, indication crowding and competitive positioning.
DAPFAM patent documents, a curated sample of US patents in Parquet with text for retrieval and classification tasks (CC BY-NC-SA 4.0). Scopus of claim text for building patent-similarity and technology-mapping models.
Sustainability information organized into general and industry-focused ESG categories with three layers, 334 subjects, and over 700 supporting indicators.
Dattito's clinical-trials-data is a Parquet-format dataset of roughly 24 million clinical trial records sourced from public trial registries. It aggregates identifiers, conditions, and standardized medical terminology into a single table for large-scale analysis.
Cardholder spending information from a non-traditional panel, broken down by geography and merchant category.
View source information and access options.
edgar_xbrl_companyfacts is a Hugging Face dataset published by DenyTranDFW that aggregates U.S. SEC EDGAR XBRL company-facts filings into a single Parquet file. The training split contains roughly 125 million rows of structured financial facts tagged by reporting period and unit.
DerivedFunction01's sec-filings-snippets is a public-records dataset of short text excerpts drawn from U.S. SEC filings, released on Hugging Face in Parquet format. It is sized for language-model fill-mask work and is not positioned as a commercial alternative-data product.
airline-disruption-data is a tabular dataset of U.S. domestic flight records published on the datasets hub, tagged as covering between 10 million and 100 million rows in Parquet format. The schema describes per-flight scheduling, delay, cancellation, and routing fields.
The airline-disruption-data-phase1b dataset is a tabular release hosted by publisher Dev123Hug456Face that compiles scheduled and actual flight-level operations data with delay, cancellation and diversion flags. It is distributed as a single Parquet split of roughly 12 million rows.