Characteristics of Most Frequent Spanish Verb-Noun Combinations

摘要

We study most frequent Spanish verb-noun combinations retrieved from the Spanish Web Corpus. We present the statistics of these combinations and analyze the degree of cohesiveness of their components. For the verb-noun combinations which turned out to be collocations, we determined their semantics in the form of lexical functions. We also observed what word senses are most typical for polysemous words in the verb-noun combinations under study and determined the level of generalization which characterizes the semantics of words in the combinations, that is, at what level of the hyperonymy-hyponymy tree they are located. The data collected by us can be used in various applications of natural language processing, especially, in predictive models in which most frequent cases are taken into account.