This was a group project for CS 206: Exploring Computational Journalism. The goal of this project is to monitor Twitter and build a dashboard that displays what key influencers (politicians, cabinet members, political commentators, presidential candidates, topic experts, government agencies, activists, etc.) have been saying on a range of political and social issues. This will be done both at the national (i.e., US) and regional (e.g., California) level. Possibly, also local (e.g., Bay Area). By issues we mean persistent or long term topics that relate to policy and can influence legislation and elections (e.g., immigration, healthcare, taxation, trade, energy, income inequality, free speech, foreign affairs, drugs, guns, mass shootings, gay rights, abortion, women’s rights, religious freedom, environment). The team will use the Twitter API to monitor a curated (and expanding) set of political and social influencers. In parallel, they will identify a variety of issues to track and build classifiers to recognize tweets that match them. For each issue they will generate a stream that can be refined in various ways -- by subtopic, influencer group, time, etc. The content of the tweets will be further analyzed and used to summarize the conversation, grouping similar expressions and URL references together. They will try to identify rising memes in the content stream and flag new developments/subtopics, key voices, highly cited articles, hashtags, etc. They will look at retweets and run searches to discover new voices to add to the pool of influencers being monitored.
Twitter has become the de facto platform for politicians and political influencers to express their views. Our goal is to create an issue-centric view of Twitter -- with a dashboard that isolates key voices of interest and then filters for key issues. This would be a useful tool for journalists, to efficiently monitor the political conversation in social media - to understand trends, the evolving positions of individuals, and to discover new stories.
My main contributions to this project were in regards to the project timeline, compiling information and data on the influencers and issues we wanted to highlight, and conducting user interviews with national, local, and college journalists to better understand how this tool can be helpful for them.