Identifying pornographic text online is practically useful to protect children from access to such adult materials. We approach to the problem as a document-level porn identification task by locating and integrating sentence-level evidence.
All the related resources (datasets and codes) will be publicly available.