Search with exclusion in parallel texts

The paper examines the method of search with exclusion in parallel texts. The described method is based on the approach that takes a text as an ordered set of wordforms. In the context of this approach, the possibilities of bilingual search by exact form, by lemma, and by morphological features are considered. This approach provides a basis not only for the abovementioned types of search, but also for search with exclusion, i. e., a search allowing to find in parallel texts such pairs of text fragments that contain a certain wordform in language A but do not contain any wordform from a given set in language B. To illustrate this idea, an example of searching fragments with implicit logical-semantic relations in parallel texts stored in a database is given. If the required wordform in language A marks a logical-semantic relation and the set of wordforms in language B contains the maximum number of variants for the translation of the required wordform into language B, the search can yield the pairs of fragments where the logical-semantic relation is expressed in the fragment in language A by the required wordform but it is implicit in the fragment in language B.