Automated Content Analysis (ACA) refers to a collection of techniques used to automatically analyze data. Because our discipline often works with texts, the term Computational Text Analysis Methods (CTAM) is also often used. Yet, while communication scholars often work with text, there is an increased focus on applying computational methods to the analysis of images as these take on a predominant role on social media.
Given that more and more communication happens online and is available in a digital format, it makes sense to analyze this huge amount of data in an automated fashion. Situated in the framework of Computational Social Science, we use techniques from disciplines like computer science and computational linguistics, including dictionary- and word-frequency based methods, natural language processing, supervised and unsupervised machine learning, and start experimenting with deep learning.
The following projects are using/studying this method:
- A cross-platform, multi-model investigation into political moral appeals
- Moved to comment
- Facts to you, opinions to me
- The comment relevance detector project
- Understanding climate polarisation and depolarisation dynamics
- Validity of computational attitude and attitude strength measures for social media data
- Personality and susceptibility to political microtargeting
- Modeling the Temporal Dynamics of the Deliberative Quality in Online Debate