From the Ripple Effect of Ideas to Machines Coloring Pictures: News from Associate Professor Jordan Boyd-Graber

From the Ripple Effect of Ideas to Machines Coloring Pictures: News from Associate Professor Jordan Boyd-Graber

Associate Professor Jordan Boyd-Graber and Yuening Hu, Ph.D. 2014, had a paper accepted to the Proceedings of the National Academy of Sciences on "Measuring discursive influence across scholarship" with collaborators from the University of Chicago and Columbia University. They explore what aspects of a piece of research (topic, researchers, institution) are most likely to change the way other scientists do science. https://news.uchicago.edu/article/2018/03/16/new-model-reveals-forgotten...

Associate Professor Jordan Boyd-Graber's advisee Alison Smith was honored with a best student paper honorable mention for "User-Centered Design and Evaluation of a Human-in-the-Loop Topic Modeling System" at the 23rd International Conference on Intelligent User Interfaces for work in collaboration with the University of Washington and Brigham Young University.  She developed a human-centered evaluation of how humans and computers can best work together to navigate large document collections.
http://www.umiacs.umd.edu/~jbg/docs/2018_iui_itm.pdf

Associate Professor Jordan Boyd-Graber had a paper accepted to the North American Association of Computational Linguistics on "Lessons from the Bible on Modern Topics: Multilingual Topic Model Evaluation on Low-Resource Languages" with collaborators from the University of Colorado.  They use the Bible (available in nearly every language) to help computers make sense of document collections written in languages that don't have software or dictionaries written for them.

Mohit Iyyer, Ph.D. 2017, Varun Manjunatha (CS Ph.D. student), Associate Professor Jordan Boyd-Graber, and Professor Larry Davis had a paper accepted to the North American Association of Computational Linguistics on "Learning to Color from Language", which uses descriptions of images to teach computers to color black and white images.

Associate Professor Jordan Boyd-Graber had an essay published in the Winter Bridge on Frontiers of Engineering published by the National Academy of Engineering, "Humans and Computers Working Together to Measure Machine Learning Interpretability", which argues that Researchers should evaluate artificial intelligence based on how well it can cooperate and communicate with humans. https://www.nae.edu/Publications/Bridge/176887/177006.aspx

A team of students led by Professor Jordan Boyd-Graber organized a human-computer question answering competition. Teams from four countries submitted systems to take on a team of top trivia players in Long Beach, CA. A computer system from Japan won both the computer competition and the exhibition match against the human team. http://hcqa.boydgraber.org