Student research opportunities
Media Geolocation and Google Maps Visualization
Project Code: CECS_147
This project is available at the following levels:
Honours, Summer Scholar, Masters
Keywords:
Visual Interfaces; Visual Summarization; Text Mining; Machine Learning
Supervisor:
Dr Scott SannerOutline:
We often browse media as a ranked list of articles. But news happens in places, blogs refer to places, and tweets come from users in places. Wouldn't it be great if we had a way to geolocate news, blogs, and tweets and an interactive visual way to display them on a map?
Unfortunately, location names are often ambiguous (Melbourne, Australia vs. Melbourne, Florida) so one needs to leverage context to accurately perform geolocation. And once this is done, it is important to figure out how to visually display this content on a map in a way that effectively imparts the necessary information to the user.
Goals of this project
In this project you will work on ways to leverage context from news, blogs, and tweets to perform geolocation; techniques range from text mining and machine learning to rule-based inference. Following successful completion of this task, you will use the Google Maps API to display this information interactively within Google Maps.
Requirements/Prerequisites
Good Java coding skills.
Links
Scott Sanner's web pageBrief overview of a blog geolocation approach
