News outlets and other content creation organizations compete for viewership on traditional channels and increasingly compete for audience engagement across multiple social media platforms. With platforms like Facebook, Instagram, Twitter, YouTube, TikTok, Snapchat, and Reddit transforming how news is consumed, understanding the dynamics of user engagement becomes crucial.
An analysis of approximately 3 million news postings and 50 million associated comments shed light on engagement patterns through views, likes, comments, and cross-platform sharing. This large-scale user engagement analysis offers an unprecedented look into how different sentiments and topics resonate with audiences, influencing the level and type of engagement a news post receives.
One key finding of the analysis is sentiment’s significant role in engagement. Positive and negative sentiments of news postings affect the volume and type of interaction, such as the potential for a post to be shared or commented on. This insight is critical for news outlets aiming to optimize their content for maximum engagement, suggesting a nuanced approach to sentiment in news topics could foster more profound interactions with the audience.
Moreover, the analysis highlighted the importance of topic selection and presentation across different platforms. A news article topic’s engagement varies widely between platforms like Facebook, Instagram, and Twitter, pointing to the unique audience preferences and behaviors inherent to each.
By combining language features and metadata features, the engagement levels of news social media posts can be predicted using advanced modeling techniques. The language features include the Term Frequency-Inverse Document Frequency (TF-IDF) features vector. The metadata features include topical, sentiment, and other general content features (e.g., content length by characters).
Aldous, K, An, J, and Jansen, B. J. (2022) What Really Matters?: Characterizing and Predicting User Engagement of News Postings Using Multiple Platforms, Sentiments, and Topics. Behaviour & Information Technology. 42:5, 545-568, https://doi.org/10.1080/0144929X.2022.2030798