News reports about the pandemic provide an enormously rich data set for communication research: How was the pandemic communicated? How did that communication change over time? And now that the pandemic seems to be coming to an end, what can we learn about public communication about illness and health?
Over the past two years, we have been able to build a line of research in which the use of metaphor in Covid communication is central. For example, we have created a large corpus of all Covid-related articles in Dutch newspapers. We have also developed and optimized a tool with the help of machine learning that can automatically identify metaphors in the corpus. This tool now performs as well as, or even slightly better than, the state of the art in the field of automatic metaphor identification - partly because we were able to use additional manual corrections/additions to the automatic coding in optimizing the tool.
Our metaphor identification tool is now performing at such a level that we can make it more widely available. In this way we contribute to the FAIR principles for open science and other researchers can also use our corpus and our tools. This requires the creation of an online environment in which the corpus is made accessible and in which the tools are integrated. In addition, user-friendly manuals must be made available to enable researchers to work independently with the data.
In addition to making the corpus and tools available, we also want to take the next step in exploring and optimizing automatic source and target domain identification.
