About M3

In the dynamic world of online media and algorithmically curated media environments, understanding the intricate online content consumption and user interaction is crucial. M3 offers a nuanced resource to social science in which it provides an up-to-date analysis of textual online media content as seen through regularly updated patterns of online media use.

M3 is divided into three core pillars:

The content pillar is dedicated to collecting and analyzing media content. From running through news outlets to scraping social media platforms, M3 aggregates and analyzes textual data to furnish a comprehensive view of the prevailing media landscape. Here, the focus is to provide as much information on the content as possible to researchers while not storing actual full texts due to legal limitations.
The use pillar seeks to categorize media-use patterns of the broader population. Based on survey as well as tracking data, the nuances of user behavior, preferences, and consumption patterns serve as guidelines to collect media content. Incorporating into the M3 database, the platform can offer a combinatory view into what a particular group of a population likely saw when moving online.
The encounter pillar builds the vital connection between the content and its users. Thinking of this as a given user’s exposure to a given content, this pillar allows to draw a descriptive picture of what media content particular media-use patterns yield. This data is compiled through the use of emulated digital agents, thus allowing to also systematically vary usage times, devices, or engagement time.

Given this background, the project puts a lot of effort into addressing the recurring call for social-scientific research infrastructure. In that, M3 is built from scratch for a variety of use cases, such as …

creating regular monitoring reports on who sees what (in the German online media sphere)
using M3 as a database for online media content that is being used
including M3 meta data from modern language models to base own research on
building on M3 to compile a sample of relevant article URLs for more specific research questions
relying on M3 as an early detection/warning system for when a particular media-use pattern is exposed to particular content
pulling up the open-source M3 code/repository to really understand how the system works under the hood
… well, what are you using M3 for?

Team

Prof. Dr. Mario HaimLMU Munich

Professor and Chair of Communication Studies, especially Computational Communication Research, at LMU Munich.