Our system allows you to find the significant values of an unknown element in a sequence of 2 or 3 words.
- To perform a simple search, type the known word in the corresponding – first or second – text box, then define part of speech of the word that is looked for OR JUST click . By default, the input is taken as a token. If you need to search for lemma, choose . In this case the input should be a dictionary form, e.g. стол, идти, красивый.
- Wait until the system calculates which words and grammatical features are most frequently used with the given word.
- To specify two words and search the third one (for example, не до X), click . The third text box will appear. Type each word in its own box leaving the box for unknown word empty and click .
- As an additional option, you may select a corpus that you want to work on:
- RNC — disambiguated subcorpus of the Russian National Corpus (5 Millions of words; is used by default).
- I-Ru — automatically annotated Internet corpus (140 Millions of words).
Note that the results for these two corpora may differ due to the size and annotation schema differences.
- You may also restrict your search by certain morphological features. For instance:
- define the part of speech for the known word (in order to distinguish between, e.g., the noun and verb печь “stove/bake”)
- define the case of a word that is looked for (e.g. accusative for the words after the preposition в “in”).
To choose morphological features click under the corresponding text box. The detailed information about annotation schemata can be found on the corpora websites:
RNC: http://www.ruscorpora.ru/en/corpora-morph.html
I-RU: http://smlc12.leeds.ac.uk/itweb/help/Russian_Language_Tutorial.pdf
- As the result you will get not all the possible words and morphological features for the given pattern, but rather the most relevant ones — when learning Russian language you should pay attention to them. These lists may be further extended, however only the most stable and frequent words and morphological features are shown. In the list of collocations with you can switch between the list of tokens (e.g., рано + вставать/овдовела/встает) and lemmas (рано + вставать/умереть/ложиться).