Abstract
In this poster, we show the under heard voices of those whose primary media is radio, the proliferation of large trend analysis over fine-grained exploratory tools for large scale media data, and the absence of exploratory tools for hyper local media like local radio. In response, we present RadioSilenced, a web-based tool that enables the semi-serendipitous (not algorithmically curated) exploration and display of talk radio call-in data to encourage a deeper understanding of the 10 percent of US communities whose primary news consumption is radio, in a whole-system view but sample-by-sample exploration closer to ethnographic observation than a more reductive, big data analysis. To our knowledge, this is one of the first large scale analysis of ASR bias in talk radio, one of the oldest forms of media.