I need to implement n-gram language model to calculate information content for semantic similarity. I found some corpus like AQUAINT-2 and NICIR-8. But these are not freely available. 

More Goutam Majumder's questions See All
Similar questions and discussions