Automatic restoration of diacritics based on word n-grams for Slovak texts

摘要

In the past and even now, many people still write texts without diacritics, especially in chat messages, e-mails or discussion posts. This issue evolved from historical reasons when people had a problem with text encoding in messages or wanted to write them faster. In this paper, we propose an algorithm based on word n-grams (contiguous sequence of n words) that restore diacritics of text written in the Slovak language. We also compare and evaluate our results with existing algorithms developed for Slovak texts.