Bot comments Report January 2024
The dataset of toxic bot comments detected in January 2024 contained over 2,000 English messages (n=2,062) posted between January 1 and January 31, 2024. The overall average toxicity score was relatively low at 0.20. However, a significant spike in toxicity was observed on January 6, 2024. This peak may be linked to an event on that day when Russian forces attacked Ukraine in Pokrovsk Raion, resulting in the death of 11 civilians (see a full list of events here). Despite the low average toxicity, some messages included highly problematic content featuring toxic terms like zionist scum, kill them, jew infestation, degenerate, ugly, moron, scum, fucking idiot, white racist and anti-white.
Around five messages were classified as highly toxic, with a toxicity score exceeding 0.8 (on a scale of 0 to 1, where 1 indicates maximum toxicity). Approximately 10% of the messages contained content categorized as Threatening, while about 2% included Disinformation. The most frequently occurring combination of categories was Politics + Religion (click here for details on category definitions).
The word cloud below displays the 50 most frequently used keywords in toxic bot comments during January 2024. The word size reflects usage frequency relative to other keywords, with terms like black, death, jewish, feminism, trump, jews were most frequently used.
The table below highlights the various toxic categories detected in the dataset, along with their respective frequencies (represented by black dots and further explained below). In January 2024, the most prevalent categories were Racism, Politics and Threat. However, the highest average toxicity scores were observed in content associated with Contempt (0.30) and content specifically targeting Jews (0.35).
The bar chart below illustrates that Profanity was the most prominent category in January 2024’s dataset of toxic bot comments, followed by Politics and Contempt. A significant portion of the toxic content was categorized under Politics, indicating that many bot comments were primarily aimed at influencing political discussions (see the explanation of the categories here).
These are the number of toxic messages by category in January 2024
The keyword network below highlights the primary focus of toxic bot comments within this dataset. Different colors represent clusters of words frequently used together. The network reveals a strong focus on Israel and Jews, along with content involving racism, where terms like blacks and whites were commonly used.
These were the keywords used by bots in January 2024 and the connections between these words