Student research opportunities
Verbose patent classification
Project Code: CECS_910
This project is available at the following levels:
Honours, Summer Scholar
Please note that this project is only for undergraduate students.
Keywords:
Natural Language Processing, Automatic Summarization, Information Accessibility, Patent Mining, Tweets.
Supervisor:
Dr Gabriela FerraroOutline:
Patents documents are assigned at least one classification code indicating the subject to which the invention relates. There are hundreds of patent class codes and thousands of subclasses encoded in alphanumeric series that nobody can remember. Usually, a patent is under more than one classification, thus understanding the interaction between those classes is even harder.
Goals of this project
The aim of this project is to apply Natural Language Processing techniques for the development of tools that makes the patent classification codes accessible to the users. First, it is necessary to map each code to its definition and then, to provide a short text, as a tweed, that summaries the classes to which an invention belongs to.
Requirements/Prerequisites
Good coding skills in Java.
Background Literature
Check out this survey about Automatic Summarization from Das and Martins, 2007.
Summarization Survey
Visit The Lens Project site
