Towards 100M Morphologically Annotated Corpus of Tajik

The paper presents a work in progress: building morphologically annotated corpus of Tajik language of the size more than 100 million tokens. The corpus is and will be by far the largest available computer corpus of Tajik: even its current size is almost 85 million tokens. Because the available text sources are rather scarce, to achieve the goal also the texts of a lower quality have to be included. This short paper briefly reviews the current state of the corpus and analyzer, discusses problems with either “normalization” or at least categorization of low quality texts and finally also the perspectives for the nearest future.