Applying AI Techniques to Patent Automation - INT-009

Preferred Disciplines: Data science, Computing science, Mathematics and Statistics, (Masters, PhD or Post-Doc) 
Project length: 4-12 months (1-2 units)
Approx. start date: May 2019
Location: San Mateo, California, USA
No. of Positions: 1-2
Preferences: Any university is fine as long as the intern can travel to Silicon Valley during the development period.
Company: PatentPal

About Company:

PatentPal is a Silicon Valley startup founded by two Harvard alumni with expertise in both patent law and artificial intelligence. We are developing AI software to automatically generate and file patent applications. Patent drafting is a $13B industry in the US and $61B worldwide. The technology we develop further has widespread applications in legal document automation.

Summary of Project:

We want to apply AI techniques, specifically, search, ML, and NLP, to generate patent applications. To do this, we want to leverage existing patent data, namely, 6.5 million patent references containing detailed descriptions on every technology ever invented. The project would involve working with our team to deconstruct each section of the patent application and design computational strategies for how to automate the mechanical and repetitive aspects of drafting those sections of the patent. For instance, some of these techniques would involve adapting existing patent data from our proprietary patent database to incorporate them into new patent applications.

Research Objectives/Sub-Objectives:

  • Design data structures for computationally representing patent data
  • Apply data science techniques to parse, clean, and enrich a database of patent data for AI applications
  • Designing search algorithms for mining the patent database
  • Design NLP techniques for understanding and generating patent data

Methodology:

  • Symbolic logic
  • Search
  • Data science
  • Machine learning
  • Natural language processing

Expertise and Skills Needed:

  • Familiarity with Python and Python libraries
  • Familiarity with search algorithms
  • Experience with implementing ML, and NLP models

For more info or to apply to this applied research position, please

  1. Check your eligibility and find more information about open projects.
  2. Interested students need to get the approval from their supervisor and send their CV along with a link to their supervisor’s university webpage by applying through the webform.
Program: