FURI | Spring 2020

Connecting the Dots: Towards Automated Dataset and Visualization Recommendation from News Articles

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

In this project, the team is investigating the efficacy of automatically recommending data visualizations to accompany a news story, even when there is no explicit data provided. To do this, the team is creating a data pipeline that combines natural language and machine learning operations, including document summarization, keyword extraction, dataset search and retrieval, and visualization recommendation. A series of user studies will be conducted to fine-tune the parameters of the pipeline, to evaluate how “best” to recommend data visualizations for text documents.


Student researcher

Shashank Ginjpalli

Shashank Ginjpalli

Computer science

Hometown: Cupertino, California, United States

Graduation date: Spring 2021