Guide to Patent Search
What is a Patent?
patent: [pat-nt or for 10, 12–15, peyt-; especially British peyt-nt] noun
- the exclusive right granted by a government to an inventor to manufacture, use, or sell an invention for a certain number of years.
- an invention or process protected by this right.
- an official document conferring such a right (aka letters patent).
- the instrument by which the government of the United States conveys the legal fee-simple title to public land.
Why are Patents So Pivotal?
Patents are useful for competitive analysis and technology trend analysis…Patents have always been analyzed in R and D project management to assess competitive position and to avoid infringement. Patent analysis is also a valuable approach that uses patent data to derive information about a particular industry or technology used in forecasting.” -Forecasting emerging technologies: Use of bibliometrics and patent analysis
The analysis of patents yields many results: the ability to spot trends, to minimize litigation, to guide research, to build portfolios, all in an effort to make better business decisions. If you are struggling with the purchase of an Intellectual Property (IP) Analysis Software program or service, or are reevaluating what you have currently, you are not alone. The world may be getting smaller, but the amount of patents available only gets larger…and the sources of data along with it. How can you be confident that your data is consistently current and complete? Are you positive that your patent analysis incorporates the mise en scène? Most importantly, are you able to perform these duties individually or do you need help? Organizations large and small the world over use professional IP analysis services and/or platforms to support their patent analysis, in a variety of ways.
Why do we search for Patents?
Patent searches performed in the 16th Century or the 21st Century have always shared three aspects:
- To find seminal patents and prior art, thereby determining whether an idea or invention is unique.
- To discover who owns a particular patent or patents in a field of interest.
- To find key inventors in order to validate (or invalidate) the particular patent searched.
To Build Your Document Library
Your patent search data is built on previously filed patents, patent-related literature, and other non-patent literature (NPL) from jurisdictions around the world. All of these documents combine to form your patent document library, the collection of which illuminates the patent story from beginning to end, highlighting original and current owners, previous and potential litigation, and current status in the IP Lifecycle.
1) to find a specific patent or document, or 2) to conduct research into an area or areas of patent investigation.
A Navigational Search is conducted when searching for a specific patent which is known to already exist. Alternatively, one may already have a set of keywords, which may be a patent or application number, the patent title, and/or the inventor’s name.
A Research Search is conducted when researching an area of new development, assessing competitors’ holdings, or seeking investors. This type of search, therefore, is a far more extensive and time-consuming process. The keyword search will return a collection of documents on a particular subject, but no single document will provide enough data. These documents may include patent filings, prior art, metadata, and NPL from journal articles to product manuals. Even though the relevance of these documents will vary widely, the analysis can reveal important information with which one can further hone the resulting list of keywords for further searching.
What is the IP Lifecycle?
The IP Lifecycle, which consists of Patents, Copyright, and Trademark, is the series of changes in the life of an idea, and the laws which govern that idea from ideation to dissolution.
Recent History of Patent Search
Performing a patent search in the 21st Century, however, incorporates technological advances now available, as well as the evolution of the function of patent search itself:
- To perform a competitive analysis that goes beyond basic searching and actually guides better business decisions.
- To locate key buyers and sellers that identify adjacent technology and potential industry partners.
- To search for component technologies which can aid in maneuvering the whitespace more effectively, and find viable opportunities to pursue novel inventions.
- To analyze global patent status so as not to miss any area of opportunity, and identify centers of excellence.
- To demonstrate due diligence, thus saving money in research, development, and potential litigation should a patent on a product already exist.
Keep in mind that because of the 18-month process for patent publication, inventors must search prior art throughout product development. These searches may reveal terms to narrow future keyword searches, but lack of prior art is not necessarily proof of novelty.
- To uncover patent litigation which can highlight the general market stage of the patent or patents.
- To determine the future direction of a given field, which can shed light on how heavy or thin the competition will be within that field.
How do we perform Patent Search?
Keyword Patent Search
Performing an effective keyword search is best enabled by becoming intimate with the language of patents. Accordingly, IP experts say the best way to learn patent searching is to read hundreds of patents. By knowing how patents are written, one can create conventional keyword searches to find patents germane to the desired technology. As searches return more pertinent results, more time can be allotted to scan and read those that are most on point.
How Much Time to Allow for Keyword Patent Search
Regardless of the type of search, navigational or research, a variety of methods will need to be included to perform a keyword search that is complete and thorough. Therefore, the need to build in appropriate time to complete the search becomes critical, bearing in mind that even expert searchers allot at least one-third of their work week to this exact task. That’s 13 hours and countless, readily available sources (though other experts say the process can take as many as 4 days) and most individuals do not have professional source access. Either way, this is an undertaking requiring more time than one might initially think.
Keyword searches return results for exact matches only. The challenge lies in brainstorming words or combination of words that will return the most relevant patents. So how to go about finding a broader range of keywords?
- By reviewing competitors’ patents for synonyms and technical terms.
- By consulting corporation subject matter experts (SMEs).
- By searching NPL.
- By examining the USPTO database of classification codes.
The USPTO offers a 7-Step Strategy Guide, including how to brainstorm terms, in order to generate a keyword list. Answering those questions will help build a list full of brainstormed words and various keywords.
Searching by Classification Codes
When using keywords to search the U.S. Patent index, an alphabetical list of classes and subclasses will result, each of which has a number or code. By searching these codes, newer and previously unthought-of relevant patents will appear using different codes.
Incorporating Boolean Search Terms
Upon the completion of several iterative keyword searches, it becomes clear which words return relevant results and which don’t. The search is then improved upon by adding Boolean operators to exclude those throwaway results. For example, if searching for the term “plane,” a search string might read plane NOT travel if searching for patents related to CAD, or better yet, plane NOT travel NOT carpentry to eliminate any hits for hand tools.
Most people are familiar with some Boolean terms (AND, NOT, OR, and quotations marks), but unless patent searches are conducted regularly, it likely won’t be clear how to 1) use proximity operators that allow for search within a set number of words, 2) search for words within a specified field, or 3) search for stem words that have different endings (air, airplane, airfield, airbag, airport, etc.). Once Boolean terms are learned, it will be important to invest time experimenting with searches to practice and ensure that the terms used are used properly so as to uncover the patents being searched.
Semantic Patent Search
The conventional search method for patent search is similar trying to find a needle in a haystack. If a keyword search is winnowing through the haystack with a pitchfork and bare hands (painstaking and painful!), then a semantic search is raking the hay over a wide area to thin it out, and then sweeping a large magnet above it. Prepare to find more needles, in fewer hours, and with considerably less repetition.
Revolutionary semantic searches find relevant patent information based on the concepts behind a normal language description. Algorithms examine word frequency, sequencing, and patterns regardless of the specific words used in the search to determine the context. Because patent authors typically use words consistently throughout a patent filing, semantic searching is highly effective at finding relevant patent documents that might use slightly different terms. Consequently, semantic searches can find patents disguised by the vague language, different terms, changes in ownership, and evolution of industry lexicon.
Better yet, semantic searching requires little training. This type of search is more intuitive because there is no need to build lengthy search strings to immediately narrow the search. Simply write a search string much in the same way one would explain a subject in conversation: with a description of the invention (which wouldn’t work in most patent search tools). Enter a paragraph describing a concept and let the patent search and analysis tool do the work.
Finally, whereas with keyword searches as much as 10+ hours are spent crafting a search to produce a full 8.5 x 11 page (or more) of related search strings, semantic search results are produced with just 4–5 words, including operators. From these results, one can then refine the search to look for classification codes, metadata, and other information. More time left over means more time for analysis of actual search results.
True Semantic Search: Be aware that there are hacked versions of semantic searches. These rely on keyword synonyms rather than algorithms. Since these keyword searches in disguise rely on keyword dictionaries, they have the same drawbacks as any other keyword search, including the inability to decipher vague language or uncover hidden terminology. In short, they lack the ability to search based on context. Even more patents will be missed if such dictionaries are not updated as a technology’s lexicon continues to develop.
Where do we perform Patent Search?
Regardless of your focus within the IP Lifecycle, the ability to search and analyze patents is key to the success of your organization. There are many use cases in which a sophisticated IP search and analysis tool can bring value for individuals, power users, law firms, and companies. Having the right information and ability to communicate insights is what separates a simple patent search from savvy IP analysis that drives business forward. As not all patent analysis tools are created equal, knowing which types of analytics are available and which will best fit your company’s goals is a critical starting point.
- Free/low-cost patent search tools – these provide basic search on raw data and will not include data corrections or the ability to see a company’s current portfolio. There will be limited language consideration, no semantic searching, narrow geographic and jurisdictional coverage, and no way to rank patents by quality or importance. Free patent search tools often have holes in the database, security and privacy risks, and as much as an 18-month lag in updated listings due to priority settings. Bottom line here is that if you choose this option…proceed with eyes wide-open on the limitations of the data.
- Services firms (law firms and consulting firms) – they perform custom IP analysis projects for a fixed price or hourly fee, where the output is a report followed by a discussion around the specific analysis requested. This is an important component when you are interested in leveraging a firms’ expertise about patent information, analysis techniques, and – sometimes – their legal advice regarding your specific technology domain.
- Patent analytics platforms – these are built on an integrated database of patents, and include data corrections, patent transactions (purchases/sales) to see current owner, status updates (e.g., have maintenance fees been paid), and often other information like patent litigation to see who is active in asserting patents. IP analysis platforms provide visualizations and analytics to discover insights and communicate them to stakeholders.