About spanish collocations on linguatools

We have compiled 4.000.000 spanish collocations from a 750.000.000 word text corpus. Each collocation occured at least two times in the lemmatized corpus.

For detecting spanish collocation in the corpus, we tagged the whole corpus with freeLing and extracted typical sequences of words such as "noun adjective" or "verb noun" and others.


Explanations on the collocations table:

In the third column there is the "tag order" of the found collocation. The tags in particular are:

VM = verb
DI = indefinite article
DA = definite article
NC = common noun
NP = proper noun
AQ = adjective
RG = adverb
SPS = preposition

"hits" in the fifth column refers to the hits of the collocation lemma in the second column.
In the second column appears the lemma of the collocation wordform as it was tagged by freeLing. 
For the sake of completeness we specify in the fourth column the type of relation between the words in the collocation, and in the sixth column the significance of co-occurrence of the constituents of the collocation.

Approximately 30.000 collocations we put online - arranged alphabetically. Within a page we sorted the collocations by tag order. 

Do you have more questions about this work? You can contact petra.prochazkova@linguatools.org in english, spanish, german or czech.