Evaluating supervised topic models in the presence of {OCR} errors

Daniel David Walker
Eric K. Ringger
Kevin D. Seppi