The strengths and weaknesses of third-party patent databases

The Daniel Keys Moran quote “You can have data without information, but you cannot have information without data” is the perfect apothegm on the significance of data. There are many databases compiling data from a variety of sources and collated in a way that users can access, interpret, search and use the information.

For the IP industry, third-party databases possess a great deal of information including patent and non-patent literature in the form of suitable descriptions and classifications. Several patent databases provide access to patent analysis with their unique capabilities. Most patent databases include a tremendous number of documents grouped by millions of inventions with millions of images from more than 100 patent authorities. A third-party patent database is a primary pillar to conducting most IP-related tasks, including assessing patentability or freedom to operate, detecting and managing infringement risk, and analysing the competitive landscape. Such databases are in particularly high demand among researchers, companies and KPOs thanks to their quick access to data analytics.

Patent databases such as PatBase and Derwent Innovation offer global patent data in English or national/regional languages from more than 60 patent authorities. These databases are amenable to extensive searching by text, classification, assignee or inventor names, and dates. Further, these databases can provide access to enhanced titles, abstracts and citation data, such as those provided by the Derwent World Patents Index™ (DWPI) and the Derwent Patents Citation Index™. Other useful features include the conversion of non-English text into English, the clustering of patent documents under patent families, legal status information on patents, maximised coverage of data and timely updates. Along with paid databases, many patent offices (eg, the USPTO) also offer various online searching tools to identify patent documents in their repository.

Databases without robust searching capability have no value; thus, database providers should offer extensive searching capabilities with a user-friendly interface to search the requisite data from various angles. An ideal database will offer to build more complex and powerful queries to find patent documents based on keywords, classes, citations and legal status. Databases should also provide advanced features such as semantic and corporate tree searching.

Aside from searching capability, other factors that are also considered when selecting the appropriate database for performing a particular task include subscription cost, maintenance service, upgrades, data download limits and response time. No single database is customised for all desired results; therefore, there are dedicated databases available for specific purposes. For example, searching design patents is different from keyword-based searching of utility patent documents; thus, databases offering searches of design patents may be desirable for only some users. Such databases work on Locarno Classification, US design or other design classification-based indexing. The databases are developed for the unique nature of data (eg, the MARPAT database for searching Markush patents through the STN platform and the USGENE database for searching genetic sequences disclosed in US patent documents). Manual analysis by technical experts is also accommodated in some databases. Scopus provides the manual categorisation of published articles and Derwent offers manually written technical summaries, abstracts and titles of patents. This combination of human and machine work makes analysis easier for users and thereby creates a more cost-effective and time-efficient process.

Several platforms are also designed to access multiple databases through one window. For example, STN provides integrated access to the most current and complete collection of global disclosed patent and non-patent, scientific and technical content. Such a platform offers access to databases belonging to various organisations (eg, chemistry content from the Chemical Abstracts Service and patent content from the DWPI).

Aside from paid database services, a few free databases are also available to search patent and non-patent documents. Google Patents is a worldwide database that is easy to navigate, readily available and free to use, and which provides a full-text translated patent search engine for patents from various countries, including the United States, Europe and Australia. The search is performed by employing variables such as search terms, date, assignee and inventor, and offers similar patent searches as well. Google Patents also enables the user to isolate the figures of a search result if they wish to focus on the figures of various patents. The patent search interface is available in multiple languages, including Chinese, English and French, with full text searching in English, French and German. Patent Cooperation Treaty applications can be accessed in Chinese, English, French, German, Japanese, Korean, Russian and Spanish. Interestingly, the free-to-use database Patent Lens offers remarkable features that enable users to explore more than 80 million DNA and protein sequences disclosed in patent documents. Some other free-to-use databases that are highly regarded by users across the globe are those of the USPTO, Espacenet and PatentScope.

Many companies have begun to offer a variety of dashboards in order to provide easy and digestible data visualisation, data interpretation and other data analytics that require minimal or no manual effort for searching and analysis. Dashboard-facilitating services can offer prior art search reports and landscape clustering based on AI or machine learning. Some of these dashboards offer citation maps providing detailed visualisations of cited patents, an accurate view of patent strength in a given domain, the relationships between inventions by organisations worldwide in order to locate whitespace, and monitoring services. For example, Derwent Innovation claims to support patent prosecution filings as well as patent monetisation and licensing. PatSeer claims to have the unique capability of multi-dimensional analysis using informative charts, custom fields, hierarchical categorisation, citation and family trees, co-citation analysis and much more. These services can help researchers and companies to gauge an idea of the scope of a domain or invention in a short time frame.

Third-party patent databases are useful tools for analysing patent and non-patent literature. In this competitive age, several organisations are offering unique services to target multiple problems. However, these databases still require significant changes to increase their desirability. Database interfaces can be improved to prepare automatic searching and clustering of patent or non-patent data in order to derive meaningful insights. In this way, they would meet the demand for foolproof automation, which would in turn provide opportunities to create various verticals in the IP industry.