Distinguished Professor Emeritus Bruce Croft, together with CIIR doctoral students Daniel Cohen and Hamed Zamani, were selected to receive a 2018 Bloomberg Data Science Research Grant Program Award.
The CIIR researchers’ work, Neural Information Retrieval with Limited Data, is one of five research projects selected for funding by Bloomberg in the 2018 international competition.
Their research will focus on studying how to design and train machine learning algorithms for information retrieval when there is no large-scale data in hand, such as in the financial domain. Developing effective approaches to learn from limited data has the additional potential benefit that search tools will be able to be effectively deployed more quickly to a new domain.
To address training IR models with limited data, they will explore weak supervision in neural information retrieval (IR) models and knowledge transfer from a domain where enough training data is available to a target domain in which limited data prevents researchers from directly training effective models.
The Bloomberg Award follows another recent industry collaboration grant on neural models for IR. Croft and doctoral student Qingyao Ai received funding this year from Amazon for their project, Representation Learning for Product Search. Croft, Ai, and Hamed Zamani met with the Amazon A9 researchers in California in March to discuss the CIIR team’s research plans.