Analysing user identity via time-sensitive semantic edit distance (t-SED): a case study of Russian trolls on Twitter

Abstract

In the digital era, individuals are increasingly profiled and grouped based on the traces they leave behind in online social networks such as Twitter and Facebook. In this paper we develop and evaluate a novel text analysis approach for studying user identity and social roles by redefining identity as a sequence of timestamped items (e.g. tweet texts). We operationalise this idea by developing a novel text distance metric, the time-sensitive semantic edit distance (t-SED), which accounts for the temporal context across multiple traces. To evaluate this method we undertake a case study of Russian online-troll activity within US political discourse. The novel metric allows us to classify the social roles of trolls based on their traces, in this case tweets, into one of the predefined categories left-leaning, right-leaning, and news feed. We show the effectiveness of the t-SED metric to measure the similarities between tweets while accounting for the temporal context, and we use novel data visualisation techniques and qualitative analysis to uncover new empirical insights into Russian troll activity that have not been identified in previous work. Additionally, we highlight a connection with the field of Actor-Network Theory and the related hypotheses of Gabriel Tarde, and we discuss how social sequence analysis using t-SED may provide new avenues for tackling a long-

Detecting Abnormal Social Network Accounts with Hurst of Interest Distribution

“…Activity features refer to a user's behavior patterns, such as active time, frequency of information published, and common clients [3,25,26]. Xin et al divided users' social behaviors into two categories: extroversive behavior features such as activity sequence and introversive behavior features such as request latency [27].…”

How did Russian and Iranian trolls’ disinformation toward Canadian issues diverge and converge?

“…Finally, my earlier work revealed the anti-Trump tone of Iranian trolls (Al-Rawi 2019a, b) in contrast to the case of Russian trolls who supported Trump and focused on divisive racial issues in the USA (see, for example, Badawy et al 2018;Kim et al 2019;Freelon et al 2020;Linvill and Warren 2020). By examining the nature of the Russian and Iranian trolls' discussion of Canadian issues, this study attempts to answer the following research questions: RQ1 What are the principal themes of Russian and Iranian trolls' tweets on Canada?…”

Russian trolls speaking Russian: Regional Twitter operations and MH17

“…The third group of studies often uses linguistic features (such as features common to native Russian speakers in English-language conversations) to detect trolls [5,11], although some studies (e.g., [16]) use a broader set of features for troll detection, including profile-, behavior-, and stop word usage-related features. Quantitative troll studies could further be divided into those taking a broad scope, and those focusing on a particular propaganda campaign, most often the 2016 US Elections [1,3,5,15,17,20] and Brexit [14,21]. This paper belongs to the first group of studies taking first a broad scope, after which focusing on the MH17 campaign.…”

“…For example, [20] identified five groups of trolls based on their behavior: right troll, left troll, news feed, hashtag gamer, and fearmonger. Finally, [17] observed that the strategic behavior of trolls changes over time.…”