Auto-Tagging of Patent Applications to Facilitate Discovery of Relevant Prior Art
Implementation Time:
24 months
Solution Provider: AI Singapore
The Intellectual Property Office of Singapore (IPOS) is the national agency responsible for the administration of IP rights in Singapore.
IPOS helps businesses use intellectual property (IP) and intangible assets (IA) to grow and is committed to building Singapore into an international hub for IP/A to drive Singapore’s future growth.
IPOS is a statutory board under the Ministry of Law.
- Upon receiving a patent application, the Patent Examiner needs to determine the relevant patent classification codes to be tagged to the application and conduct prior art search using both codes and relevant keywords in public e.g., Google Patents, Espacenet as well as subscribed commercial databases
- Effectiveness of the search is subject to the accuracy of the codes/keywords and Examiner expends a significant amount of time to manually come up with the codes, keywords and subsequently the set of relevant prior art
- The abovementioned are currently carried out manually by the Examiner
How can IPOS create a minimum viable model to (i) auto-tag patent applications with the relevant classification codes and (ii) automatically generate a set of relevant prior arts.
An Al solution was developed using natural language processing to:
- Analyse incoming patent applications
- Assign classification codes to the application
- Perform a search from an internal database
- Determine the prior arts relevant to the application
- Rank the prior art in an order of relevance
- Tag the prior arts to the application
Outcomes
- Estimated to save 25% of current man hours per search-examination work
PDF Document
This is the last content of this tab. If you do not see any resources above, it means the solution provider have not provided any resources. Feel free to contact Solution Provider for more information.
Implementation Time
24 months
Use Case Brochure
