Choosing the Right Search Type

With three distinct search types available within the NLPatent platform, it’s essential to understand their differences to ensure you make the optimal selection for your search project. Here’s a guide to help understand which search type would best suit your needs. 

As a primer, both the Patent Number and Natural Language search are AI-based search types, relying on our proprietary Large Language Model (LLM) to generate results. These search types don’t require any keywords to generate results instantly and they allow for further advanced AI functionality, such as iterative learning (Refine) and AI generated analysis (Relevance Analysis). 

Patent Number Search

Inputting a patent number as your search mechanism allows for the identification of similar patents efficiently. Our AI examines the full specification to retrieve relevant results instantly. This method is particularly effective for conducting validity assessments, streamlining the process with its precision and speed. It can also be used to quickly assess the state-of-the-art related to an existing patent document (just remember to remove or adjust the Priority Date filter, which is automatically applied for the earliest priority date). 

Natural Language Search

For patentability searches, we recommend employing the Natural Language search type. Whether you're a seasoned patent professional, an entrepreneur conducting preliminary checks, or an engineer exploring similar inventions,  Natural Language search offers simplicity and accessibility. Its strength lies in its ability to interpret inventions described in plain language, catering to a diverse range of users.

Moreover, Natural Language search isn't limited to patentability searches alone. In specific scenarios, it can outperform patent number searches for validity assessments, especially when targeting very specific claims or limitations. For example, by simply copying and pasting a claim of interest,  Natural Language search ensures that results are focused and accurate.

While conducting a natural language search, remember to ensure that you give enough context and detail so that our AI understands the scope of the invention well. We also have a full article detailing how to write a Natural Language query. 

Keyword Search 

While NLPatent supports keyword searches, it's essential to note that this method does not leverage specific AI features such as Relevance Analysis and Refinement. We recommend initiating your search journey with either Natural Language or Patent Number search, followed by applying a keyword filter to ensure your search results contain whatever keywords you specify. 

However, if you're interested in exploring patents based on specific Assignees/Owners, Applicants, or Inventors, the "Skip Query" option in keyword search is a valuable tool. By selecting this option, you can streamline your search process and obtain targeted results efficiently. Since this search type is void of advanced AI features, users can select up to 10,000 search results, unlike other search types, which are limited to a maximum 1,000 references. 

In essence, by understanding the distinctions between Patent Number, Natural Language, and Keyword searches, users can select the optimal search type for their needs. .