Large-Scale Topical Analysis of Multiple Online News Sources With Media Cloud

Authors

Jason Chuang, Sands Fish, David Larochelle, William P. Li, Rebecca Weiss

Published

January 1, 2014

Identifying topics in news, tracking their temporal dynamics, and understanding how different media sources cover them have important theoretical and practical implications for journalism researchers, producers, and consumers.

The explosive growth of online news sources, however, suggests that scalable approaches to topical analysis are needed. We introduce our ongoing efforts to enable large-scale topical analysis of the Media Cloud corpus, a repository of over 200 million online news articles.

Our initial experiments with 90 days of articles from 21 top media sources suggests that statistical topic modeling can identify reasonable news-related topics and produce interesting early insights into the online media ecosystem. We are currently examining mixed initiative approaches to automate the process of topic extraction and increase the quality of the extracted topics.

Finally, we discuss our further research directions on large scale news monitoring and measurement as well as analysis tools for news consumers and producers.

Read more of our recent research

12/12/2023

How the news talked about the Israel-Hamas conflict in its first month

7/27/2023

CUNY Community Media Explorations

Media Cloud has partnered with the City University of New York’s Center for Community Media on several different media analysis projects in recent years, producing insightful and relevant findings on community-level media ecosystems.

5/11/2023

RedditMap.Social: A novel map of Reddit based on user clustering

The University of Massachusetts Amherst part of our team, within the Initiative for Digital Public Infrastructure, set out to explore a different type of media ecosystem through mapping the social media platform Reddit. This endeavor is a part of a long-standing project, supported by the Knight First Amendment Foundation, to map the wide landscape of all social media online.