Our data pipeline is the foundation of everything we do. It begins with raw patent data from the USPTO and ends with structured datasets that are easy to explore and analyze. This process includes:

• Data ingestion from multiple USPTO sources
• Cleaning and normalization to correct errors and inconsistencies
• Schema mapping to align data with our internal structure
• Versioning and release management to ensure transparency over time

We update our data quarterly and document all changes in our release notes. This methodology ensures that users always have access to the most current and reliable data available.

Inventor and assignee names often appear in multiple formats across different patents. Our disambiguation algorithm groups together records that refer to the same person or organization, even when names are misspelled, abbreviated, or reordered.

This process includes:

• Name normalization (e.g., “J. Smith” vs. “John Smith”)
• Affiliation matching using co-inventor and assignee data
• Machine learning models trained on historical patent data

The result is a more accurate picture of who is inventing what, where, and when, which is critical for network analysis, innovation tracking, and policy research.

Patent data often includes inconsistent or incomplete location information. Our location standardization methods clean and enrich this data to support geographic analysis.

We:

• Parse and validate city, state, and country fields
• Match locations to a curated database of known places
• Resolve ambiguities (e.g., “Springfield” in multiple states)

This enables users to explore innovation trends by region, map inventor networks, and compare patenting activity across geographies.

Many patents are funded by federal grants or contracts, but this information is often buried in unstructured text. Our government interest method extracts and flags these patents using natural language processing (NLP) techniques.

We:

• Scan patent documents for statements of federal support
• Extract agency names and grant numbers
• Tag patents with structured metadata for easy filtering

This helps users identify federally funded innovation and evaluate the impact of public investment in research and development.

Patents are categorized using multiple classification systems, each with its own structure and purpose. This allows users to explore innovation by technology area, compare trends across industries, and align patent data with other economic indicators.

Our classification method maps patents to:

• International Patent Classification (IPC)
• Cooperative Patent Classification (CPC)
• U.S. Patent Classification (USPC)
• WIPO Technology Fields
• NBER Technology Groupings

We don’t just build tools, we test them. Our evaluation methods use real-world use cases and performance metrics to validate the accuracy and usability of our data.

We:

• Benchmark disambiguation accuracy
• Test API response times and reliability
• Collect user feedback through our forum

This continuous improvement process ensures that PatentsView remains a trusted resource for patent data analysis.

Download large sets of patent data in a tab-delimited format. You can choose raw, cleaned, or processed data. This format works well with most data analysis tools like R or Python. It’s great for deep research and custom analysis.

Use our application programming interface (API) to access patent data programmatically. You can create custom searches and pull data directly into your apps or websites. The API includes detailed documentation and a Swagger interface to help you get started quickly.

A user-friendly tool that offers an efficient visual interface for building custom patent datasets. The PatentsView Data Explorer allows users to:

• Enter search criteria using a simple filter interface
• Explore and refine patent data visually
• Download tailored datasets for further analysis
• Use the Insights tab to view trends and summaries
• Navigate with ease, whether you're a technical or non-technical user

Explore patent data through interactive charts and graphs. You can search by keywords, filter by location, and compare different attributes. These data visualization tools help you spot trends, patterns, and relationships in the data.

Search more than 40 years of patent data. This tool helps you find connections between inventors, patents, and companies. It’s a powerful way to explore innovation trends and discover new insights.

These guides explain how our data is organized. They include definitions for terms, field names, and attributes. Use them to better understand the structure of our datasets and ensure accurate analysis.