integrated database (e.g., OECD 2010 used PATSTAT and Li et al. 2013 used
Espacenet) provides more comprehensive results. However, Dechezlepretre et al.
(2009) and Hong et al. (2013) obtained fewer patents from the integrated database
PATSTAT than the Derwent Innovations Index database. This is likely due to use of
an incomplete search expression.
Patent family analysis is popular. A patent family is a set of similar patents taken
in various countries to protect a single invention. Therefore, analysis by patent
family more accurately reflects the total number of inventions. We extend these
previous studies through the creation of a more complete database using more
suitable search expressions to obtain data from the Derwent Innovations Index
database. We highlight developmental stages over a 40-year period and analyze
characteristics by country.
8.1.3 Patent Database and Retrieval Strategy
The Derwent Innovations Index (DII) database of the Web of Science (from the
I.S.I. Web of Knowledge) was used to search and analyze patent data. The DII
database is a widely accepted source of patent data covering over 14.3 million basic
inventions from 40 patent-issuing authorities worldwide. In patent bibliometrics, the
most fundamental task is to set the range of retrieval and choose appropriate index
words. Based on previous studies, search indicators include two dimensions: search
topics and International Patent Classification (IPC) codes.
We collected keywords of types of carbon capture processes, combined as single
topic search expressions using logical operators. For more rigorous search results,
we used a thesaurus to find synonyms of IPC codes related to those keywords and
combined them as single IPC code search expressions using logical operators. The
IPC system is the most popular hierarchical classification of patents among countries
or organizations with official patent offices and was launched by the World Intellectual Property Organization (WIPO). We tested dozens of combined search topics
and IPC codes, comparing results of inputting different expressions, and then
adopted the most suitable expressions. In testing, we found a dozen globally
renowned vehicle companies. However, they do not operate professionally in the
domain of carbon capture. We found most patents from vehicle companies were
related to pollution control of automobile exhaust gases. Therefore, we excluded this
technology by using the logical operator “not.” This appears to be the first time that
exhaust control technologies have been excluded from patent data searches using
logical operators, resulting in a dataset that is more concentrated on carbon capture
patents. The expressions of search topics and IPC codes used are summarized in
Tables 8.2 and 8.3, respectively.
The final patent data search was conducted in April 2016. A total of 9847 patent
files were found, filed over a time span of 40 years from 1976 to 2015. For each
patent in our database, several data fields were extracted, including authorization
year, assignee name, assignee code, and IPC code.
8 Spatial-Temporal Distribution of Carbon Capture Technology According to Patent. . .
157
Espacenet) provides more comprehensive results. However, Dechezlepretre et al.
(2009) and Hong et al. (2013) obtained fewer patents from the integrated database
PATSTAT than the Derwent Innovations Index database. This is likely due to use of
an incomplete search expression.
Patent family analysis is popular. A patent family is a set of similar patents taken
in various countries to protect a single invention. Therefore, analysis by patent
family more accurately reflects the total number of inventions. We extend these
previous studies through the creation of a more complete database using more
suitable search expressions to obtain data from the Derwent Innovations Index
database. We highlight developmental stages over a 40-year period and analyze
characteristics by country.
8.1.3 Patent Database and Retrieval Strategy
The Derwent Innovations Index (DII) database of the Web of Science (from the
I.S.I. Web of Knowledge) was used to search and analyze patent data. The DII
database is a widely accepted source of patent data covering over 14.3 million basic
inventions from 40 patent-issuing authorities worldwide. In patent bibliometrics, the
most fundamental task is to set the range of retrieval and choose appropriate index
words. Based on previous studies, search indicators include two dimensions: search
topics and International Patent Classification (IPC) codes.
We collected keywords of types of carbon capture processes, combined as single
topic search expressions using logical operators. For more rigorous search results,
we used a thesaurus to find synonyms of IPC codes related to those keywords and
combined them as single IPC code search expressions using logical operators. The
IPC system is the most popular hierarchical classification of patents among countries
or organizations with official patent offices and was launched by the World Intellectual Property Organization (WIPO). We tested dozens of combined search topics
and IPC codes, comparing results of inputting different expressions, and then
adopted the most suitable expressions. In testing, we found a dozen globally
renowned vehicle companies. However, they do not operate professionally in the
domain of carbon capture. We found most patents from vehicle companies were
related to pollution control of automobile exhaust gases. Therefore, we excluded this
technology by using the logical operator “not.” This appears to be the first time that
exhaust control technologies have been excluded from patent data searches using
logical operators, resulting in a dataset that is more concentrated on carbon capture
patents. The expressions of search topics and IPC codes used are summarized in
Tables 8.2 and 8.3, respectively.
The final patent data search was conducted in April 2016. A total of 9847 patent
files were found, filed over a time span of 40 years from 1976 to 2015. For each
patent in our database, several data fields were extracted, including authorization
year, assignee name, assignee code, and IPC code.
8 Spatial-Temporal Distribution of Carbon Capture Technology According to Patent. . .
157