What is Normalization?
Normalization is the process used in IPlytics to ensure comparability of patent indicators. Each indicator value is benchmarked against a control group of patents with the same patent office, technology classification (IPC/CPC), and publication year.
How is Normalization calculated?
For each patent, the indicator value is compared to the average value of similar patents that share:
- the same patent office
- the same IPC/CPC class
- the same publication year
For example, a patent filed in a specific year and technology field is evaluated against the average performance of patents with the same characteristics. The resulting value reflects how it performs relative to this benchmark.
How are normalized values interpreted?
In order to provide comparability of different patent portfolios, all indicators are normalized by averages of patent control groups of the same patent office, same IPC/CPC class and the same publication year. A patent from 2010 filed in the US with the IPC/CPC H04W that e.g. receives 6 foreword citations is compared to the average forward citation counts of all US/2010/H04W patents. If the forward citation count is 6 and the average is 3 the indicator would be 2, meaning twice as high as average.
Normalized indicators allow comparing patent portfolios of different age, technological focus or country. All patent indicators are normalized by 1 and can thus be interpreted as follows:
Value greater than 1 (>1)
A value greater than 1 indicates that the patent performs above the average of comparable patents in terms of technology, country, and year.
Value less than 1 (<1)
A value less than 1 indicates that the patent performs below the average of comparable patents in terms of technology, country, and year.
Why is Normalization important?
Normalization enables meaningful comparison between patent portfolios by accounting for differences in:
- age of patents
- technological field
- regional patenting practices
This ensures that indicator values reflect relative performance within a comparable context rather than absolute counts.
Click here to learn more about indicators in IPlytics.