17 Jun
2020

Efficient use of search tools for prior art and patent searching

Co-published

Potential applicants are recommended to carry out a prior art search to check whether their invention is novel. These searches can create stronger claim development, as well as verifying that the invention does not infringe another’s intellectual property, thereby avoiding any potentially expensive litigation further down the line.

When performing a prior art search, patents are the main (and best) source for identifying patent documents (non-patent literature can also be useful). A patent database search will look for particular ideas and technologies by keyword, classification, date, inventor and assignee, among other things, and can be carried out on numerous free and paid databases.

Free patent databases

Since each country carries out the patent examination procedure before granting exclusivity rights to an invention, most have developed their own search portals in order to carry out an initial patentability examination and simplify the examination process.

To further enhance the efficiency of these portals and keep up with the rapid pace of development, most countries have moved towards the amalgamation of machine learning and AI to organise patent documents (eg, into International Patent Classification (IPC) and/or Cooperative Patent Classification (CPC) categories, along with other docketing processes based on technical divisions).

The top free databases with research and analytics functions are provided by:

  • Google Patents;
  • the EPO;
  • the USPTO;
  • the Japan Patent Office;
  • the Korean Intellectual Property Office;
  • the Canadian Intellectual Property Office;
  • the China National IP Administration; and
  • IP Australia.

Third-party databases

There are several paid databases that operate on deep-learning AI and natural language processing, which provide a comprehensive guide to portfolio analysis, document comparison and high-performance searches across a large number of countries. These are equipped with machine translation to translate all non-English patents into English for easier understanding. The AI ​​behind these databases further enhances other important features, such as filtering and sorting the shifting of patented datasets. Another interesting function of these tools is the similarity search, which looks at phrases or paragraphs in the patent and showcases a list of similar patents based on the input.

The formation of a query depends on the analyst’s know-how of the concept and its iterations. These searches are built with synonyms, proximity and Boolean operators and different types of classification (eg, IPC, CPC and US and F-terms) to obtain the exhaustive set of prior art documents related to the concept disclosed in the invention.

The abovementioned features are common to most paid databases. However, the most appropriate and unique specifications of several paid third-party databases are listed below.

Orbit, Questel

This is a highly respected database to perform patent searching for prior-art searches and landscape analysis. It can access more than 54 million patent families, 100 million patents and 12 million design patents. It also provides worldwide patent coverage. It is possible to search for non-patent literature as it provides access to 108 million scientific publications, including books, research papers, journals and articles. Further, efficient resource sharing is available due to its sub-account feature. Here, multiple sub-accounts can be associated with a single primary account, which allows for the export of a shortlisted patent dataset to these sub-accounts for further analysis, while the primary account remains free for subsequent formation and running of search queries.

Derwent Innovation, Clarivate

This is an excellent tool for projects that require long and complex search queries, as it allows significant flexibility to a patent search process, especially in the life sciences and wireless sectors. Data intelligence provides a stunning graphical representation of the patent dataset for results interface based on assignees, inventors and legal states, among other things, which can come in handy in landscape projects. Further, it provides advanced filtering features such as forward and backward citations, legal status and INPADOC patent family searches.

Patseer, Gridlogics

For a service provider or an innovation-driven company, Patseer can function an optimum decentralised solution, as it allows the creation of a group where multiple members can work, customise and manage their project online. It can work as a boon in case of a landscape project, as different members can create an online taxonomy where they can add or remove patents in that category and can directly export the dataset or charts based on this later on. Sorting and filtering the relevant patents is much easier with this database, as it provides an outstanding dashboard for visualisation and analytics where it is possible to play with the results in multiple ways. There is also a filtering feature based on the number of occurrences of particular keywords used to search for a patent.

Patsnap

To be a unicorn in the field of ideas and technologies, you need to keep an eagle eye on competitors, which the insights feature of Patsnap helps to achieve. This is useful in identifying the patent value, top authorities, patent type and filing trends of any desired company. For an individual patent attorney or strategist, the most tedious task can be done with a single click by using the playbook list feature, which analyses patent value, litigation threat and litigation history, and simulates a merger and licensee locator. This database provides a quick check on whether a patent has contributed to any standard.

Comment

R&D professionals and inventors need tools to efficiently analyse the disclosures closest to their innovation from a pool of ideas presented to them. Although free tools provide easy access to every individual, they also have drawbacks, which may affect the accuracy and efficiency of the search process and potentially risk the project. These holdbacks can be corrected by including features such as Word Intelligence to automatically highlight the synonyms of keywords used in the search strings, graphical and statistical analysis of datasets, and patent sorting and filtering based on the high relevancy order. It is hoped that some of these features will be included in subsequent updates to other patent databases.

For further information contact:

Nitesh Chaurasiya
Effectual Knowledge Services Pvt Ltd
View website

Sanchit Kalra
Effectual Knowledge Services Pvt Ltd
View website

This is a co-published article whose content has not been commissioned or written by the IAM editorial team, but which has been proofed and edited to run in accordance with the IAM style guide.