Why patent classification tools are indispensable

A classic terminology problem

Suppose the target is a method for removing contaminants from a semiconductor-wafer surface with a laser. One document may say “laser cleaning of a semiconductor wafer surface.” Another may describe “laser ablation of surface contaminants in IC manufacturing.” A third may refer to “removing surface particles by pulsed laser irradiation of a substrate.”

The concepts may overlap even though the wording does not. A relevant classification can create another retrieval path across those wording differences. Still, it does not guarantee that every relevant document receives the same code. The sound approach is to use classifications, keywords, citations and related-document expansion together.

For teams that want a structured patent-search workflow, Eureka IP Search’s Novelty Search extracts technical features, builds multiple search strategies and compares close prior art feature by feature.¹ Classification review remains part of professional search judgment; an automated result is not proof of complete coverage.

IPC in patent classification tools

What is the IPC?

The International Patent Classification, administered through WIPO, is a hierarchical system of language-independent symbols used to classify patents and utility models by technical field. WIPO releases a new IPC version that enters into force on January 1 each year.² The current IPC FAQ describes almost 80,000 classification entries rather than the older figure of approximately 70,000.³

The hierarchical structure of the IPC

The IPC descends through sections, classes, subclasses, main groups and subgroups. For example, WIPO uses the following structure in its current guidance:

Main-group symbols end in /00. Subgroup hierarchy is shown by dots before titles in the official scheme; the digits alone do not establish the hierarchy. Always read the parent title, notes, references and definitions rather than treating the symbol as a standalone label.³

IPC applications in patentability searching

Advantages:

• It supplies a shared international framework across technical fields.
• The official scheme, guide, statistics and downloadable files are freely available from WIPO.
• Many patent-search systems expose IPC fields or filters.

Limitations:

• Its broad international role means that some technical areas need finer searching through CPC, FI/F-Term or database-specific tools.
• A new version enters into force annually, so saved strategies should record the version and be reviewed after revisions.
• Classification practice and data completeness can differ by authority, database, document and time.

A search expression such as IPC=B60L50/50 is only illustrative. Field names, punctuation, spacing, truncation and hierarchy behavior differ among databases. Check the chosen database’s current search help before running or documenting the query.

CPC in patent classification tools

What is the CPC?

The Cooperative Patent Classification is an extension of the IPC jointly managed by the EPO and USPTO. The EPO describes nine sections—A through H plus Y—and approximately 250,000 classification entries.⁴ CPC can therefore provide a finer route into many technical topics than IPC alone.

Do not describe CPC as having a fixed monthly update cycle. The current scheme history shows several effective versions in some years, but the dates are release-specific rather than a promise that a new version will appear every month.⁵

CPC vs. IPC comparison

The table compares the official systems, not the completeness of any particular commercial database. Coverage must be checked in the database being used.

A CPC refinement example

Within semiconductor-device thermal management, CPC H01L 23/34 covers arrangements for cooling, heating, ventilating or temperature compensation. Narrower places include subjects such as heating, cooling facilitated by device shape, cooling by air and cooling by liquid. The exact wording and available subgroups can change with scheme revisions, so use the current CPC scheme and definitions instead of copying an old code list.⁶

FI/F-Term in patent classification tools

What are FI and F-Term?

The JPO describes FI and F-Term as Japanese patent classification systems. FI is based on IPC and adds finer subdivisions for Japanese search needs. F-Term organizes patent documents through multiple technical viewpoints within defined themes. The JPO currently reports approximately 190,000 FI entries and 330,000 F-Term entries.⁷

Why Japanese searching may need FI/F-Term

FI and F-Term can improve focus when relevant Japanese documents are difficult to separate with IPC or keywords alone. They are especially useful when a searcher has identified Japanese family members, applicants, inventors or a technical theme that merits a jurisdiction-specific supplementary search. Their value comes from the additional retrieval route, not from a universal rule that every Japanese search must use them.

FI/F-Term search channels

• J-PlatPat provides free access to Japanese patent information and FI/F-Term classification lists.
• The JPO provides an IPC-FI-CPC parallel viewer for comparing corresponding scheme entries.
• Commercial platforms may expose classification helpers and classification-specific fields. For example, Patsnap Analytics supports patent-search and analysis workflows, while its Help Center documents IPC, CPC, FI and F-Term search fields.⁸

Practical example: J-PlatPat identifies F-Term theme code 2H045 as “Mechanical optical scanning systems.” It is not a generic code for all optical elements. Combine the theme with the relevant viewpoint codes and verify their current English or Japanese definitions before relying on the result.⁹

How to find the correct classification number

Method 1: start from relevant known patents

1. Run an initial keyword search with a few core technical concepts.
2. Select several genuinely relevant patent documents rather than accepting the first results automatically.
3. Review all IPC, CPC and, where relevant, FI/F-Term symbols shown in their bibliographic records.
4. Open each symbol in the current official scheme. Read its title, parent and child places, notes, references and definitions.
5. Test the candidate symbol against known relevant and irrelevant documents. A useful code should improve retrieval without hiding important neighboring concepts.

WIPO’s 2026 IPC Guide expressly recognizes reviewing the classification symbols of closely related patent documents as one way to identify an appropriate subclass.¹⁰

Method 2: use classification search and assistance tools

• WIPO IPC Publication: browse the hierarchy, definitions, notes and revision information.
• WIPO IPCCAT and STATS: use text or statistical assistance to generate candidate IPC places; treat suggestions as starting points for review.
• CPC scheme and definitions: browse current symbols and revision notices.
• USPTO Patent Public Search: use the database’s documented CPC fields and syntax.
• J-PlatPat: browse FI and F-Term themes and viewpoint codes.

If the invention description is incomplete, classification selection becomes unreliable. Eureka IP Drafting’s Invention Disclosure workflow can turn rough engineering notes into a structured disclosure and highlight missing details.¹¹ This is a separate intake task; it does not decide the legally correct classification or replace classification searching.

Method 3: build a reviewed field library

Experienced searchers often maintain reviewed starting symbols for recurring fields, such as lithium batteries, wireless communications, pharmaceutical compositions or genetic engineering. Treat these as starting points, not permanent conventions. Before reuse, confirm the current scheme version, inspect revision notices and test whether the invention falls into a neighboring or cross-cutting class.

Strategies for combining classification searching and keyword searching

Basic principle

Use classification symbols to define one or more technical neighborhoods, then use keywords to focus on the claimed feature, relationship, material, sequence or effect. In parallel, keep a keyword-only path to catch documents that lack the expected classification.

The USPTO’s multi-step search guidance follows a similar sequence: begin with keywords, review close documents, expand with relevant CPC classifications, examine citations and then broaden into foreign patents and non-patent literature.¹²

Combined search query example

This is conceptual syntax only. Translate it into the field codes, quoting, truncation and Boolean rules of the chosen database. Record the exact executable query in the search log.

Iterative optimization process

1. Start with a broad IPC main group, CPC subclass or keyword set.
2. Review relevant documents and collect additional symbols, terms, citations, applicants and inventors.
3. Narrow with CPC subgroups or FI/F-Term viewpoints where the data supports them.
4. Apply feature-specific keywords inside the classification-defined set.
5. Run parallel keyword-only and citation searches to test what classification searching missed.
6. When Japanese evidence matters, inspect FI/F-Term codes and search J-PlatPat or another verified source.
7. Document databases, versions, queries, filters, dates, exclusions and stopping rules.

Common pitfalls when using patent classification tools

Pitfall 1: assuming one classification symbol is sufficient

A patent can receive multiple symbols for different technical aspects. Review the full classification set on close documents and test adjacent classes instead of relying on one code.

Pitfall 2: ignoring hierarchical relationships

A short title can hide important scope notes and references. Browse the full tree and definitions, including broader parents and narrower children, before deciding which level to search.

Pitfall 3: copying syntax between databases

Fields such as IPC=, .CPC. or other database aliases are not interchangeable. Check current help documentation and save the actual query that produced the results.

Pitfall 4: assuming CPC provides complete global coverage

CPC is detailed and widely used, but a code’s presence in the scheme does not prove that every relevant document from every authority has complete CPC data in the selected database. Combine IPC, CPC, national classifications, keywords, citations and family review as the matter requires.

Key takeaway

Patent classification tools are valuable because they create technical retrieval paths beyond terminology. IPC supports a broad international route, CPC provides finer subdivisions, and FI/F-Term adds Japan-specific search perspectives.

The strongest practice is not “classification instead of keywords.” It is a documented, iterative strategy: classification symbols define technical neighborhoods; keywords focus on the specific solution; citations, families and national tools test the remaining gaps. Qualified reviewers should assess search limits and the legal significance of any prior art found.

Sources and verification

1. Patsnap Eureka, IP Search. Accessed July 29, 2026.
2. WIPO, International Patent Classification. Accessed July 29, 2026.
3. WIPO, IPC Frequently Asked Questions. Accessed July 29, 2026.
4. EPO, Cooperative Patent Classification. Accessed July 29, 2026.
5. CPC, Scheme and Definitions Update History. Accessed July 29, 2026.
6. USPTO, Current CPC Scheme H01L. Accessed July 29, 2026.
7. JPO, Classification. Accessed July 29, 2026.
8. Patsnap Help Center, How Can I Find a Specific Classification Code?. Accessed July 29, 2026.
9. J-PlatPat, F-Term Theme Code List 2H. Accessed July 29, 2026.
10. WIPO, Guide to the IPC (2026). Accessed July 29, 2026.
11. Patsnap Eureka, IP Drafting. Accessed July 29, 2026.
12. USPTO, Multi-Step U.S. Patent Search Strategy. Accessed July 29, 2026.

This article provides general information, not legal advice. Classification data, database syntax and scheme versions change. Verify current official records, document the exact search environment and obtain qualified patent review before relying on a search conclusion.