news-topic-stress-impact | News Topics on Facebook-based Stress Impact Scores

news-topic-stress-impact is a Jupyter Notebook library. news-topic-stress-impact has no bugs, it has no vulnerabilities, it has a Permissive License and it has low support. You can download it from GitHub.

The amount of time people spend on social media for e.g., Facebook, Twitter etc. is constantly increasing. Teens now spend up to nine hours a day on social platforms, while 30% of all time spent online is now allocated to social media interaction. And the majority of that time is on mobile - 60% of social media time spent is facilitated by a mobile device. Source. With such widespread use of smartphones, the role of social networking websites has greatly transformed. Becoming popular as a fun way to share your vacation pictures with the world, Facebook feed has now turned into a one-stop destination for everything from creepily relevant advertisements to latest news on violence and destruction around the world. How many times have you logged into Facebook with the intention of checking updates from dear ones only to find unpleasant news of politics and violence on top of your news feed?. Negative news can significantly change an individual’s mood leading to distress and degradation in productivity. In this project, I analyze the potential stress impact of articles shared on Facebook and the variance of this impact based on the broad topic of the news article. Intuitively, an article on violence/war is expected to induce a more negative reaction compared to an article on sports/lifestyle. My initial goal is to validate such hypotheses using Facebook data available publicly. If we do observe significant variations based on news topics, we can potentially incorporate the topic dimension into news recommendation engines. For example, during working hours, the recommendation engine can place low priority for negative news with high stress scores if previous activity by the user for similar news articles is not found. For the data collection, I utilize Facebook API to collect information including article link, likes and "reactions" information to news articles shared on the official pages of reputed newspapers around the world. Additionally, for each news article, I find out the broad topic category of the article either from the article's link or by scraping the article's main page. With the collected data, I perform various analyses including relationship between the negative and positive reactions to an article and the variance in stress score across news topics. For those unfamiliar with Facebook reactions, these are additional emotions one can express for photos, posts and other content on Facebook apart from the "like" option which was the only reaction present previously. The available reactions are shown in the figure below the title.