SoylentNews

canopic jug [soylentnews.org] writes:

Back in November 2020, the NATO Stratgic Communications Center of Excellence in Riga, Latvia published an analysis of the coordinated online harassment of Finnish government ministers [stratcomcoe.org]. The conclusion is that the attacks and astroturfing are largely free from automated activity, aka bots. The report includes statistics, lots of analysis, and several illustrative graphs. The main topics triggering the online abuse were the SARS-CoV-2 pandemic, immigration, EU relations, and social policies. Finland is not a NATO member but the lessons learned from studying the coordinated harrassment can be generalized to the alliance.

This report is informed by the findings of three recent Finnish studies, one of which investigated the extent and effects of online hate speech against politicians while the other two studied the use of bots to influence political discourse during the 2019 Finnish parliamentary eleections. The first study released by the research branch of the Finnish govenment in Novemeber 2019, found that a third of municipal decision-makers and nearly half of all membes of Finnish Parliament have been subjected to hate speech online.

[...] As social media platforms continue to grow in political importance, so does their use as a means for engaging with and criticising individual government officials with little or no consequences. An additional aim of our study was to determine the role, if any, bot accounts play in disseminating abusive messages, and whether such bot activity displayed characteristics of coordination. Based on previous Finnish studies analysing the impact of bots during election periods, we hypothesised that we would observe low levels of automation and coordination. Our findings confirmed this theory; our algorithm attributed less than 3% of abusive messages to bot-like accounts. However, the more significant finding was that over half of abusive messages were sent by anonymous accounts. Anonymity erases accountability online. This can have the effect of emboldening users to voice their dissatisfaction with ministers through unfiltered, abusive messages. It is possible for people to operate many anonymous accounts. However, our data do not show clear patterns indicating single users sending abusive messaging from multiple fake accounts. The unfortunate conclusion is that much of the offensive, sexually explicit, expletive-filled abuse targeting government officials is written and published by individuals.

The data was collected from March 2020 through July 2020. The report defines "hate speech" early on and categorizes it into generalized or directed, implicit or explicit. Quite a bit of material is devoted to the algorithms used to collect the data and to help do the analysis. Despite taking digs at "anonymity [techdirt.com]", which is sometimes agitated against by a key NATO member [cjr.org], and including hypotheses critical of it, there was little given to support the negative view. Perhaps the term could have been defined at the outset, since it seems used in several different meanings throughout the report.

Noticeably, the algorithms for sorting and prioritization of messages within social control media are not addressed, and therefore neither is the effect the non-chronological order has upon perceptions and opinions. As a result little was mentioned about the influence excerted through social control media upon individuals and resulting in modified behavior online. Thus the report ends up mistaking social control media for communications media or platforms for public engagment rather than calling them out for being about mass manipulation of opinions and propaganda.

Previously:
(2018) Politicized Trolling is More Harmful than Fake News [soylentnews.org]
(2016) Astroturfing is Psychological Warfare [soylentnews.org]

Original Submission