Inside Trending Topic Algorithm: How Do Human Interactions Drive Public Opinion in an Artificial Environment

• Published in: Social science computer review Vol. 41; no. 1; pp. 234 - 248
• Main Authors: Russo, Vanessa, del Gobbo, Emiliano
• Format: Journal Article
• Language: English
• Published: Los Angeles, CA SAGE Publications 01.02.2023; SAGE PUBLICATIONS, INC
• Subjects: Algorithms; Artificial environments; Datasets; Network analysis; Opinion leaders; Public opinion; Social network analysis; Social networks; Tagging; Tags; algorithms; Twitter; computational social science; hubs; Python
• ISSN: 0894-4393; 1552-8286
• DOI: 10.1177/08944393211041501

The object of this research is to exploit the algorithm of Twitter’s trending topic (TT) and identify the elements capable of guiding public opinion in the Italian panorama. The underlying hypotheses that guide the whole article, confirmed by the research results, concern the existence of (a) a limited number of elements at the base of each popular hashtag with very high viral power and (b) hashtags transversal to the themes detected by the Twitter algorithm that define specific opinion polls. Through computational techniques, it was possible to extract and process data sets from six specific hashtags highlighted by TT. In a first step through social network analysis, we analyzed the hashtag semantic network to identify the hashtags transversal to the six TTs. Subsequently, we selected for each data set the contents with high sharing power and created a “potential opinion leader” index to identify users with influencer characteristics. Finally, a cross section of social actors able to guide public opinion in the Twittersphere emerged from the intersection between potentially influential users and the viral contents.