One of the premier research institutions in the world, Cornell University encourages its students “to pursue unpredicted lines of thinking in order to effect change on local and international scales.” To further the school’s vision of discovery and our own core value of fostering innovation, I’m excited to announce that Workday has awarded four academic research grants to Cornell faculty members in fields related to natural language processing, artificial intelligence, and machine learning.
The 2018 recipients of the Workday Gift for Computer Science and their proposal titles are (drumroll, please . . .):
- Yoav Artzi: Mapping Sequential Instructions to Actions for Natural Language Interfaces
- Claire Cardie: Text Summarization for Enterprise-level Document Collections
- Thorsten Joachims: Fairness in Learning to Rank
- Vitaly Shmatikov: Protecting Integrity and Privacy of Machine Learning
Each of these winners will receive a grant of $75,000 to fund their research. We are thankful to all of the applicants, whose exceptional proposals resulted in a spirited debate among our leadership team before deciding on the four finalists.
The projects ultimately selected were those that were most applicable to our business, had promise to apply proposed techniques to real use cases, had a high likelihood that their research will lead to a positive outcome over the next year, and had the potential to help us build longer-term relationships with some of the brightest minds in technology research.
We want to nurture and grow the best thinking and execution on the nascent technologies that promise to transform business and society at large.
For instance, Dr. Shmatikov’s group plans to design and implement machine learning pipelines that ensure the integrity of machine learning models while protecting the privacy of individuals whose data is being used for training purposes.
The research of Dr. Joachims and his group addresses the question of fairness in search rankings. Now that they are ranking everything from apartment rentals to job candidates, there’s a responsibility for ranking systems to be useful to the person doing the search but also “fair” to the people its results might impact.
Dr. Artzi and his group address the issue of context when uttering search queries. His research will develop new context-dependent models for mapping sequences of instructions.
Dr. Cardie’s group will take initial steps to develop methods for summarizing and understanding enterprise-level document collections. They intend to investigate methods to produce generic and role-specific summaries of regulations, policies, and procedures.
As an alumnus of Cornell University, I am especially honored to have the opportunity to work with the faculty once again, this time through Workday. On the heels of the $250 million Workday Ventures fund launch earlier this year, Workday’s collaboration with Cornell is further proof that we want to nurture and grow the best thinking and execution on the nascent technologies that promise to transform business and society at large. We believe it’s imperative to look beyond our walls, where innovation takes place throughout the collective technology universe.
Cornell’s Computing and Information Science (CIS) Dean Greg Morrisett told me, “Workday has already been such a wonderful supporter of CIS, but we are excited that they have now moved to the next level and are helping to support our faculty and graduate students doing research on a number of projects with exciting potential for Workday’s business.”
Workday will begin engaging with the recipients starting this summer, and researchers are encouraged to publish the results of their projects.