Kalev is speaking at DataTV 2020 today, alongside organizations like the BBC and the MPEG consortium.
GDELT is collaborating with the Internet Archive’s Television News Archive to securely and non-consumptively analyze more than a decade of television news coverage through Google’s Cloud Video AI API. From captioning search to language modeling, visual video search to advanced video analytics and visual misinformation research, summary airtime files to raw frame-level annotations, how do we transform tens of terabytes of machine annotations into actionable insights at scale? How can semantic entailment divide broadcasts into “stories” and how can we richly contextualize broadcasts by reverse image search of each second of airtime across the open web, enriching broadcasts with online coverage and EXIF metadata? Finally, how can we apply all of these tools to explore how television news is covering the COVID-19 pandemic?