Website Search
Find information on spaces, staff, and services.
Find information on spaces, staff, and services.
Many first encounter the task of patent searching while trying to patent an invention, but there are other reasons to search for patents. Patent searching experts typically categorize patent searching by the reasons for the search, which inform the strategies and tools used. Each type of search may use a combination of strategies, which will be explained in section 4.2, and databases, explained in 4.3. The different types of patent searches fall into two rough categories, however: patent searching performed for legal reasons and patent searching performed for informational and design reasons.
Legal patent research typically requires exhaustivity; the end goal is to produce a set of all relevant patent literature and then determine the scope of legal coverage via analysis of the claims, terms of protection, and other details. Below are some of the common patent searches performed to determine legal rights.
Note that patent applicants and others invested in determining the legal status of patents frequently employ professional patent searchers (including patent attorneys and agents) to conduct patent searches on their behalf, as experts can often find patent literature that novices cannot. Patent agents and attorneys can also offer interpretations of what is found. This lesson provides tips for preliminary patent searching only.
The goal of a patentability search is to find all prior art that might be relevant to a potentially patentable invention. The patent applicant (or patent attorney or agent) then analyzes this prior art to determine if they can create a patent application that will meet the requirements of novelty and non-obviousness. Patentability searches should cover granted patents and patent applications from all possible patent offices. To be truly thorough, the patent applicant will also search for prior art in non-patent literature using scholarly article databases and other resources. Exhaustive patentability searches are sometimes called “state-of-the-art” searches, although this term is also sometimes used to describe market analyses that may not involve patentability (see below).
Freedom to operate searches help determine whether an inventor can produce and market an invention in a specific territory without fear of infringement. Some inventors may not intend to patent their inventions due to the cost and labor of the process or a desire to keep the technology open-source. Some inventors may initially intend to patent their inventions and, during the course of a patentability search, discover that their invention is fully anticipated by prior art and likely unpatentable. Freedom to operate searches take patentability out of the equation and focus instead on avoiding infringement. This involves asking questions surrounding the status of disqualifying prior art, for instance:
Validity/invalidity searches establish whether a patent or one or more of the patent’s claims has been granted in error. The searcher attempts to find prior art that was not acknowledged during the patent process to disqualify the granted patent. An inventor might do this to establish freedom to operate or to respond to an infringement suit brought by the owner of the patent.
Most prescribed patent search strategies are geared toward those interested in the legal rights afforded by patents; some experts have identified this as a result of outreach by the USPTO and other entities most interested in fostering economic growth and suggested that these strategies may not suit all patent searchers’ needs.
Patents are an underutilized and rich information resource, however, full of technical detail and, through their citations of prior art and classification systems, well-suited to following trends in history, business, and technology. Searching patent literature for informational purposes does not require exhaustivity nor as formalized an approach as a patentability, freedom to operate, or validity/invalidity search might.
Engineers and others engaged in product design can use patent literature to inform their processes. As patent literature includes new and detailed information not available in popular and scholarly sources, it can help designers:
At the University of Wisconsin-Madison, where faculty, students, and staff are obligated to disclose new inventions (as detailed in Lesson 3 of the Introduction to Intellectual Property micro-course), having a general knowledge of patents in a discipline can also provide a baseline from which to assess whether you have an “invention” to be disclosed.
Although patent searching during the product design process goes generally unacknowledged as a distinct type of patent search, using a patentability search or freedom to operate search as a model for this type of search may prove unnecessarily cumbersome. Instead, these searches may benefit from improvisation, creativity, and using databases with simple, flexible functionality. Searchers might initially borrow strategies from legal-focused searches, but find they can search more efficiently for the information needed without such a structured approach.
Needs for historical patent searching vary widely, but might involve identifying patent records for a known invention or patents invented by a specific individual or owned by a specific company. The information at hand may determine the best database and strategy in performing the search. Some factors that might influence a historical patent search include:
State-of-the-art and landscape searches can help you understand a specific technology or market. These searches combine patent searching on a broad area of technology with accompanying searches of trade and scholarly journals, as well as industry and market research. Goals of these searches include gaining insight into the developed and as-yet-undeveloped technology in an industry, and identifying competition and market peers. Some alternately use the term “state-of-the-art searching” to refer to an exhaustive patentability search that incorporates non-patent literature (see above).
Regardless of the type of patent search you intend to perform, you will likely need to use a strategy that combines keyword searching with other types of searches, possibly in multiple patent searching databases. The first part of this section explains some basic components you can use in a search strategy. This is followed by some combined search strategies that make use of these components.
Most patent databases allow for keyword searching, although the scope, abilities, and quirks of these searches vary widely.
For instance, Google Patents is based on Google’s traditional search interface, where users start by entering keywords in a single box. Google Patents then searches a huge database of scanned patent documents from most international patent offices (in some cases back to 1791) for records in which all of the words occur anywhere within the document. This typically yields lots of results, many of which may not be relevant.
On the other hand, the quick search in the USPTO PatFT database allows you to search specific or all fields of U.S. granted patents back to 1976, but limits you to two words or phrases at a time (the advanced search for USPTO allows for more flexibility, but requires more familiarity with Boolean operators, nested expressions, and familiarity with USPTO fields). This offers extreme specificity, but is significantly harder to use.
As discussed in Lesson 3 regarding titles of patents, patent applicants may not use intuitive language, colloquialisms, and brand names in either their titles or descriptions. For keyword searching, instead consider the following:
Due to differing terminologies, different database coverages, and other factors, even expert keyword searchers will likely miss patents relevant to their search unless they follow up with additional strategies.
One productive way to begin a search is to identify a given researcher with expertise in a type of technology, or to identify the maker of a product that closely resembles what you intend to design or patent, and to search for either the inventor name or applicant, assignee, or owner name in the appropriate field.
For some historical searches, this might be the only step required to identify the desired patent(s). For other searches, this may lead to terminology for keyword searches, cited references to explore, and classification codes to expand your search.
Once you have discovered at least one relevant granted patent or patent application, you can expand your search using citations to other patent literature. Most patent searching databases include hyperlinks to patents referenced in the patent application or granted patent (with some caveats; for instance, the USPTO PatFT and AppFT databases only link to other U.S. patents referenced, not foreign patents). Many patent databases, including Lens.org, Google Patents, and Espacenet, also link from patent documents to later patent applications and granted patents that referenced them as prior art.
As explained in Lesson 3, patent examiners assign classification codes to patent documents so that similar inventions are grouped together even if the inventors used different terminology when filing their applications.
Cooperative Patent Classification (CPC) is a single system used by the U.S. Patent and Trademark Office and the European Patent Office (EPO) to classify patents and is compatible with the International Patent Classification system used more broadly. Examiners typically assign multiple CPC codes to a single patent.
Each CPC can be broken down by levels of specificity. For example:
A63F 9/16
Section A (HEALTH; AMUSEMENT)
Class 63 (SPORTS; GAMES; AMUSEMENTS)
Subclass F (CARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR)
Group 9 (Games not otherwise provided for)
Main group 16 (Spinning-top games)
Augmenting a keyword search with CPC searching may seem like extra work, but patent examiners make heavy use of these codes in their prior art searching, which means that learning to use CPC codes in the patent search process can better help you navigate the literature on their terms. Understanding and using CPC is therefore essential for patentability, freedom to operate, and validity/invalidity searches, but can also helpfully define the scope of a technology for a state of the art or landscape search, and can also unearth invaluable patent literature for the design process.
One way to get to know the CPC system is to experiment with Espacenet’s Classification Search feature, which allows you to search on keywords within the classification definitions and to navigate through the levels of classification by clicking headings, starting with lettered Sections at the top level. An advantage of viewing CPC codes in this way is that you can see neighboring classifications, which may turn out to be just as relevant or even more relevant for your search.
As mentioned above, most prescribed search strategies are conceived with the idea of establishing legality–they emphasize a systematic, consistent approach geared toward comprehensiveness. Patent researchers engaged in less formalized product design, state-of-the-art/landscape, and historical patent searching may find that harvesting these strategies for parts may be more productive than following them strictly.
The USPTO offers a 7-step strategy to identify all prior art in U.S. granted patents and patent applications. This strategy requires developing keywords to search a CPC scheme hosted on the USPTO website, then using relevant CPC codes to run searches in the CPC fields of the USPTO PatFT and AppFT databases (described below), which include U.S. patents and patent applications, respectively. As acknowledged in the official strategy document, this is not a truly comprehensive prior art search in that it locates only U.S. patent information rather than all prior art.
A patent librarian at the University of Maine produced a variation on the USPTO’s recommended strategy that substitutes the more robust classification search on the EPO’s Espacenet database for the USPTO’s CPC scheme. Although it uses the EPO’s classification search, this is still focused on U.S. patent literature. (A patent searcher could, however, expand the search to international patents and patent applications by checking a box next to the desired classification(s) in the Espacenet classification search.)
Although both of the previous strategies begin with brainstorming keywords, they recommend using these keywords to search classification schemes directly. As classification scheme definitions are developed around consistent terminology, this will only yield good results for searchers who luck into using the same terminology as the creators of the scheme.
A way around this is to use brainstormed keywords to search patent literature directly by running searches in databases with full-text access to lots of patents. This may be particularly helpful for concepts and technologies that are too new or specific to have entire CPC classifications dedicated to them. Once the searcher has found records for a few relevant patents, they can use the CPC codes assigned to those patents to expand the search to include granted patents and patent applications in which different terminology is used to address the same functionality.
For instance, you might try one of the following methods:
In all cases, the key to patent searching is repeating the process, trying new keywords, new classifications, and combinations and, typically, using a combination of patent searching databases.
No single patent searching database does everything well. The databases created by the USPTO, EPO, and other patent offices have the most definitive information for those offices; however, third-party databases offer unique search features, user-friendly interfaces, and even full-text searching on patent literature that the official patent office databases may not. Each patent searching website may visualize the information associated with a patent document differently (and may include updated information on details that may change after issuance, such as assignee), but most include downloadable PDFs of the original patent document.
Below is a list of useful patent-searching databases and a brief summary of their content and capabilities as of 2019. Although there are subscription patent searching databases available, these are all free to the public (with some conditions, in the case of PubEAST and PubWEST).
The USPTO website includes several patent searching databases. These offer definite, official information on U.S. patents and patent applications; however, each database is dedicated to a particular type of patent document (i.e., granted patent or patent application) or aspect of patent protection (i.e., the application process or patent assignee status and history).
Espacenet is the official EPO patent search database and includes patent literature from patent offices throughout the world, including the U.S.
There are a number of third-party patent searching websites developed by corporations and non-profit entities. Two of the most user-friendly and useful are Lens.org and Google Patents.
PubEAST and PubWEST are public-facing versions of patent searching tools used by USPTO patent examiners. Unlike the databases above, which are all freely available on the web, PubEAST and PubWEST are only available at USPTO Public Search Facilities and by appointment at designated Patent & Trademark Resource Center libraries (including Steenbock Library at UW–Madison).
As U.S. design patents and plant patents do not map directly to patents in other countries and regions, searchers for these types of patents may require different strategies.
U.S. design patents are searchable on the USPTO websites. The USPTO does not classify design patents using the CPC system, but rather continues to classify them under the older U.S. classification system. This system presents a good starting point for searchers. One recommended strategy is as follows:
U.S. design patents are also available in Google Patents and Lens.org and will be included in search results for keyword searches. Searchers can limit their results to U.S. design patents on Google Patents by selecting “Design” as the Type. Searchers can limit their results to international design rights (including U.S. design patents) by selecting Design Right as the Doc Type. Lens.org also includes U.S. Classification as a filter, which can be used to limit to specific U.S. design patent classification codes.
Searchers can also use the WIPO Global Design Database, which includes U.S. design patents, as well as comparable industrial design rights from other countries.
Like design patents, U.S. plant patents are searchable on the USPTO website. As is not the case with design patents, the USPTO classifies plant patents using both the CPC system and the older U.S. classification system. Using the U.S. classification system may yield better results for plant patent searchers.
As mentioned in Lesson 3, U.S. plant patents often contain color photographs in the Drawing Sheets section in order to best represent the appearance of the plant. These are only available in black and white via the USPTO website and most other patent databases; however, current full-color print versions are made available at Patent & Trademark Resource Centers, including Steenbock Library. In addition, the University of Maryland Library and New York Public Library have scanned and made available online many full-color plant patent images.
You had intended to patent and market your product in the U.S. only, but, while searching for patents in Espacenet, you identified a Japanese patent that anticipates your planned claims in full. What kind of search can you perform to see if marketing in the U.S. is still a legal option?
In discovering a Japanese patent that anticipates your claims, you have effectively established that patentability in the U.S. is unlikely, as this patent counts as prior art and would likely lead to the USPTO rejecting your application on the basis of novelty. You have effectively already conducted a patentability search. Landscape searches and historical searches are focused on identifying trends in an industry or locating specific historical patents, respectively, and are not primarily concerned with legal rights.
In discovering a Japanese patent that anticipates your claims, you have effectively established that patentability in the U.S. is unlikely, as this patent counts as prior art and would likely lead to the USPTO rejecting your application on the basis of novelty. You have effectively already conducted a patentability search. Landscape searches and historical searches are focused on identifying trends in an industry or locating specific historical patents, respectively, and are not primarily concerned with legal rights.