Anurag Bhardwaj Faisal Farooq Huaigu Cao Venu Govindaraju Topic based language models for OCR correction. 107-112 2008 AND http://doi.acm.org/10.1145/1390749.1390766 conf/sigir/2008and db/conf/sigir/and2008.html#BhardwajFCG08