Researchers at the University of Michigan School of Information (UMSI) earned awards for best paper and best poster at the 2019 Web Conference in San Francisco.

Qiaozhu Mei, associate professor of Information, took home a best paper award at the conference for his paper “Emoji-Powered Representation Learning for Cross-Lingual Sentiment Classification.” The paper describes the development of a machine learning tool that uses emoji to classify online sentiment across all languages. 

Zhenpeng Chen, Sheng Shen, Ziniu Hu, Xuan Lu and Xuanzhe Liu are co-authors on the paper. 

Assistant professor David Jurgens won best poster at the Web Conference with “Demographic Inference and Representative Population Estimates from Multilingual Social Media Data.” The research describes a method to accommodate and adjust for sampling bias in social media data across all languages.

Zijian Wang (formerly at UMSI), Scott Hale, David Ifeoluwa Adelani, Przemyslaw

Grabowicz, Timo Hartman and Fabian Flöck are co-authors. 

The annual Web Conference, held since 1994, is considered the premiere forum for researchers, policy makers, developers and users to discuss the Web and its ramifications. Ten UMSI researchers presented papers at this year’s conference. 

- UMSI News Service