Art Of Using Classifications To Add Precision In The Patent Search
Patent Search & Analytics Prior Art Search

Art Of Using Classifications To Add Precision In The Patent Search


A prior art search can be conducted with various strategies, one of the most effective ways of searching for the promising prior art is by using patent classifications. Patent classification is a system of patent offices to categorize patent applications, according to the technical features of their content. Cooperative Patent Classification (CPC), International Patent Classification (IPC), United States Patent Classification (USPC) and File forming term (F-Term) are some of the most prominent patent classification systems. The patent classification system makes it easier to retrieve patent documents relating to a similar field of technology even if different documents use different terminology to describe the same feature.

Patent classification

Patent classification is a system developed by Patent Offices to define invention based on their technical field in a hierarchical form. Generally, a patent is assigned with one or more classifications based on the technological content present in the specification. Since the patent office has already done the work of classifying the patents for you, one can use the patent classifications to conduct a précised search to unleash all the promising prior arts that could have not encountered while conducting a keyword-based search.

Various Classification Systems

Different classification systems used for searching prior art are as follow:

1) International Patent Classification (IPC): One of the most précised and important classification systems is the International Patent Classification (IPC), developed and maintained by the World Intellectual Property Organization (WIPO). The IPC system is subdivided into sections, classes, subclasses, groups, and subgroups. The IPC classification currently divides technology into around 70,000 sub-areas. In the IPC, technologies are divided into eight sections with approximately 70,0000 subdivisions. The IPC classification is assigned by the national as well as regional intellectual property offices that publish the patent documents.

2) Cooperative Patent Classification (CPC): The Cooperative Patent Classification (CPC) is an augmented version of the International Patent Classification (IPC) and is jointly developed & managed by the United States Patent and Trademark Office and European Patent Office. It is divided into nine sections, A-H and Y (Y is an addition to the sections present in the IPC system i.e. A-H), which in further sub-divided into classes, sub-classes, groups, and sub-groups. There are approximately 250000 classification entries currently present in the CPC system.

3) United States Patent Classification (USPC): United States patent classification is an antiquated patent classification system, it was used by the United States Patent and Trademark Office (USPTO) to classify US patent applications. Patent documents before 2014 are searchable using USPC codes as USPC was later on succeeded by the Cooperative Patent Classification (CPC).

4) File Forming Term (F-Term): File forming term is a system to classify the Japanese patent documents according to the technical features described therein. F-terms are precise and prove useful while conducting prior art searches to find Japanese prior arts.

5) Locarno Classification: Locarno classification is an international classification system for classifying industrial designs. These classifications are used to effectively search for design patents/registrations.

    Identifying the Relevant Patent Classifications

    > There are several ways to identify relevant patent classifications; some of them are as follow:

    > Using the Classifications given in subject patent

    > Extracting the Classification used by examiner in prosecution history

    > By extracting classifications from the closet prior arts identified during keyword search

    > By identifying relevant CPC classes using keywords directly through Espacenet and further browsing through the class tree.

    > Extracting classifications from the citations of subject patent

      Why classification based searching is important?

      While conducting a prior art search none of the known methods should be underestimated. You never know which method will give you the desired prior art. There are several reasons why classification based searching is important for prior art search, some of them are as follow:

      > The classification based searching adds precision to a search, as it narrows the search area. 

      > Keyword-based searching can be tricky for example the word “valve” can mean either a mechanical valve or an electrical valve. Hence, classifications related to the specific field of technology must be used to fetch relevant results that could have been left out during the keyword-based search. 

      > Classification based searching is less ambiguous as compared to the keyword-based searching. 

      > Searching with keywords can often produce inaccurate and incomplete results due to the language in which patent documents are written and the terms used therein.

      A Real Scenario in Which Classifications Search Helps Us To find 102 prior art

      While working on an invalidation project related to mechanical, we have noticed that in the patent very general terms are used such as vacuum, insulator, valve, etc. So, initially, we have made some narrow strings and used all search strategies. But will not able to unleash the promising result. At last, we have made a very broad string to ensure that there is no prior art left. We have only used the important keywords of the subject patent without even using their synonyms. Still, we faced a problem that the number of patents to analyzed is very high. There was no concept which we can apply to restrict our strings. So, we left with only one option i.e. to use a relevant class to restrict the string. Although, we have already made strings using relevant narrow classes but got no promising art using them. So, this time we have chosen the class very carefully and use a broad class which was covering the classes used by the subject patent as well as by examiner in the prosecution history. Using that class with a broad string, the number of hits is reduced up to a great extent. Luckily, we found a promising reference in that string which was published in some other jurisdiction. It was exactly using the concept of the subject patent. This is how classification search not only to add precision in our search but also helps us to found 102 prior art.


      Every searching technique has its significance; so does the classification based searching. One searching technique cannot become a substitute for the other. When performing searches one need to obtain results that are as accurate as possible and classification system is one of the strategies which can be used to add precision to the search so that accurate result can be obtained in desired time. However, the selection of classifications should be done very carefully. A searcher should not rely on a single search strategy and should try to use all possible search strategies.

      Authored By: Shubham Chaudhary, Research Associate, Operations Team

      Leave a Reply

      Your email address will not be published.