We have recently released the second edition of IPGOD 2018 and a new data product IPLORD.
Intellectual Property Government Open Data (IPGOD) is published annually by IP Australia and includes over 100 years of registry data on intellectual property (IP) rights administered by the agency. It also contains derived information about the applicants who filed these IP rights.
Produced by IP Australia’s Office of the Chief Economist, IPGOD 2018 is the most significant release since the first in 2013 and has been published in two editions. In the first edition released in May 2018, we used a machine learning process to construct the applicant identifier, ipa_id, in order to better match and consolidate name variations of the same applicant. This was done to overcome multiple names that were discovered for the same applicant. We also improved our algorithm to identify whether an applicant is a corporate entity or a private individual.
We took this process further in the second edition and consolidated around 11,000 ipa_ids for Australian business applicants that also changed names. This was facilitated by our improved ABN matching technique to correct around 18,000 (11%) of about 160,000 existing ABNs in the 2017 data as well as identify over 40,000 new ABNs in the latest year’s data.
Using the better quality ABNs in commercially obtained business-level data allowed us to identify 6 455 large firms in IPGOD 2018 and 6 407 in IPGOD 2017 instead of 4 000 previously, which represents a significant 60 per cent increase in correctly identified large firms engaged in IP activities. This information was then used to derive the type of applicant, which includes manually identified universities and research organisations.
We also added geocodes to applicant addresses and made around 50 000 corrections (23 per cent) to 217 000 existing addresses and mapped nearly 20 000 new addresses. We used this information to derive local government areas, Australian Bureau of Statistics (ABS) Statistical Area Level 2 and greater capital cities as well as federal electorates.
Finally, IPGOD 2018 has a new table called IPGOD100, which is a consolidated patent summary to help answer queries about patents. It currently includes 43 application-level data items, using data from 14 separate IPGOD tables.
These enhancements to IPGOD make it a significantly improved data set, with the potential to help researchers better understand IP activity in Australia over time.
Alongside IPGOD 2018 we have released a new data product named Intellectual Property Longitudinal Research Data, or IPLORD for short. It is constructed around IP rights applicants rather than applications and tracks the IP activity and profile of each applicant repeatedly over time, generating longitudinal data on the stocks and flows of applicant-level IP rights.
IPLORD was built from the Intellectual Property Government Open Data (IPGOD). Where IPGOD captures the life-cycle of individual IP rights applications, IPLORD systematically transforms the application information in IPGOD into derived variables to track each applicant’s IP activity over time.
In its current release, IPLORD contains 68 data items to include key details about an applicant as well as its IP activity and profile over time, with further disaggregation by technology fields in patents and trade mark types in trade marks. Key details include an applicant identifier, type and geographic origin of applicants as well as the Australian Business Number (ABN) for Australian business applicants and a variable to indicate whether the business in question is alive or no longer existing.
There are two versions of IPLORD. The public version is available on data.gov.au, the Australian Government’s open data platform. A subset of this, excluding private and international applicants, was created for the Australian Bureau of Statistics (ABS) for its integration into the Business Longitudinal Analysis Data Environment (BLADE).
At its core, BLADE contains annual tax records for all actively trading Australian businesses. By linking business-level financial and IP data, BLADE will offer the opportunity to investigate how IP activity, business performance, competition and innovation may be interrelated.