Rewriting the Future of Work: Professor Emily Liu
Leading businesses looking to make sense of the digital revolution often approach Stevens Institute of Technology for insight.
School of Business professor Emily Liu is often the one who provides it.
More than a decade as a researcher with IBM gave Liu the ability to move seamlessly between projects. The company had rotated her from finance to healthcare to education to other areas.
“Every year, I had to learn something from scratch,” recalls Liu. “It’s a good way to learn about different industries as you complete different projects and deliver results to customers.”
As she gained more experience, Liu became more interested in data science and artificial intelligence, particularly with regard to their applications to finance and the life sciences. That’s what ultimately led her to Stevens and its business school, where she takes a focused approach to research into the blockchain, deep learning and text mining — with the support of students eager to play a role in new work.
“The students at Stevens are really smart; they enjoy being hands-on with data; they are very passionate about technology,” says Liu. “They’re very ambitious. When they run into a problem they cannot solve, they ask the right questions and, with some guidance, are able to come up with solutions.”
Student breakthroughs occur frequently in her courses, such as one on web mining that forms part of the university’s graduate Business Intelligence & Analytics program.
Liu and her students also interface with industry. One team recently took on a complex project for Genesis Research, a life sciences consulting company seeking a machine-learning approach to literature reviews for research projects. Literature reviews are integral to Genesis’ work, but time consuming: the average review involves researchers examining thousands of articles, often for periods of more than a year. Improving the process allows those researchers more time and resources to focus on actual problem-solving.
Alongside three of her students — two of whom had served as interns at Genesis — Liu developed a new deep-learning framework to facilitate data extraction once articles included in a literature review have been identified. That framework also leverages a so-called few-shot learning model to efficiently discover patterns during text analysis and better train neural networks to identify useful articles in a data set. Liu’s team will continue working with Genesis to further refine the model, she says.
“Considering the small set of data, the performance of the model was quite decent, but there is some room for us to improve with additional testing,” notes Liu.
Her work is another example of the powerful approach that sets Stevens’ business research apart from the pack. By discovering new ways for machines, data technologies and human experts to work together, the university’s researchers are helping rewrite the future of work in ways that complement individual strengths while also improving efficiency.