The Artificial Intelligence Opinion Monitor (MeMo:KI) is monitoring:

1. The formation of public opinion by means of a monthly population survey, the results of which can be found in our dashboard. Besides an overview of the results, you can also find there data sorted according to, for example, age group.
2. The opinions published on the topic of artificial intelligence by means of an analysis of media coverage: we look every month at the most popular print and online media, and evaluate which topics are reported in connection with AI.
3. The networking of actors in Twitter discussions: we evaluate every six months which organizations and individuals are using Twitter to talk about AI. This page provides full analyses of Twitter communication.

The following applies to all Figures: connections are created through the function of the so-called @-Mention (this means that accounts are directly mentioned or tagged) or a retweet. Figure 1 provides a dense network to explore, and illustrates those accounts mentioned by at least 20 different accounts in the six-month period in tweets containing the terms “AI” or “Artificial Intelligence”.

Why is this important? Restricting the number has allowed us to filter out such accounts that others deem to be less important. We do this especially for reasons of clarity. This means that the graphs only show a small section of the actors that are addressed on the topic of AI.

The color of the accounts in the graphs indicates the type of actor involved, e.g. whether it is a political account (e.g. belonging to a political party) or a journalistic account. However, there can also be individual persons behind an account, these tweeting privately and withholding information about their professional or occupational function. These have been anonymized for reasons of data protection. However, there are also accounts of people who do not express themselves privately, but are understood as public figures, these remaining in the network graphs if other accounts mention them frequently. What does the network depict?

What is represented in the network?

A network graph depicts two elements: nodes and edges. In Figures 1 and 2, the colored dots represent Twitter accounts, or “nodes”. The larger a node is, the more often that account was mentioned or retweeted with an @-Mention. Instances of different accounts referring to each other are depicted by lines between the nodes, or “edges”. The wider such an edge is, the more often the accounts refer to each other. Besides the size of the nodes and the width of the edges, what is also significant is the position in the network: the more often an account is referred to by other accounts depicted, or itself refers to these accounts, the more central this account is in the network. However, if an account that is referred to frequently (large node) has a more peripheral place in the network, then this indicates that the references come from accounts that are not depicted in the network.

Network of accounts mentioned by at least 20 different accounts on the topic of AI

To make things a little clearer, we have reduced the number of accounts depicted in the following Figures. The network graphs that follow are less dense than Figures 1 and 2 above. The Figures on the left (3 and 5) show the current network structures from the second half of 2021; next to them are Figures 4 and 6, which show the network structures from the first half of 2021.

Network of accounts mentioned by at least 30 different accounts on the topic of AI

The comparison of the two half-years in Figures 5 and 6 is even clearer, but also less complex. Here, only those accounts are illustrated that were mentioned by at least 50 different accounts in Twitter communication about AI.

Network of accounts mentioned by at least 50 different accounts on the topic of AI

Who tweeted the most? Which accounts were mentioned most often? Who referred to others most often?

Besides the network graphs, the following Tables show which accounts posted the highest number of tweets per six months, which actors were referred to most often, and which actors referred to others most often. It is important to note that we have again anonymized private individuals here for reasons of privacy. But they were not removed from the Table because this would mean losing important information for the network: private actors are a relevant group for discussions on the topic of AI.

Do you have further questions about our analyses and evaluations? If so, our research team comprising Kimon Kieslich and Pero Došenović at the University of Düsseldorf will be happy to help. All scripts for data collection and analysis are available upon request from kimon.kieslich@hhu.de. Data protection does not allow us to share Twitter data. Dr. Esther Laukötter at CAIS is the contact person for journalists and for matters of science communication.