You can’t suggest that?!

In this article, we study correction of spelling errors, specifically on how the errors are made and can model them computationally in order to fix them.The article describes two different approaches generating suggestions for three Uralic languages: Estonian, North Sámi South Sámi.The first approach modelling is rule-based, where experts write rules that describe kind made, these compiled into finite-state automaton models errors.The second data-based, show a machine learning algorithm corpus humans have it creates neural network errors.Both require collection error corpora understanding its contents; therefore also actual seen detail.We find while both create systems, with current resources expert-build systems still more reliable.