dlex

The Annotated type bigrams table

The Annotated type bigrams table lists all (different) sequences of two annotated types identified from the annotated corpus.

In this table, frequencies for bigrams are differentiated with respect to the part of speech tags given to these bigrams during annotation. For example, for the bigram that has the orthographic form die Demokratie, the table contains two rows with two different bigram frequencies: one for die/ART analyzed as an article, and one for die/PRELS analyzed as a relative pronoun (like in diejenigen, die Demokratie wollen).

For each of the two components of a bigram, you can query any variable that is also available for a single annotated type (like in the Annotated types table). A possible query on the Annotated type bigrams table could, for example, deliver the set of all bigrams consisting of der followed by an adjective that has a low frequency, but scores high in regularity (e.g., der vertikalen, der verlangten, der verhaßten).

The Annotated type bigrams table offers the following filters or variables for output:

• Surface filters
  • Type bigram
• Frequency filters
  • Annotated type bigram frequency
• Conditional probability
  • Annotated type bigram cond. prob.

Additionally, for each of the two components of the bigram you can select any of the filters available in the Annotated types table:

• Annotated types