Extracción automática de colocaciones terminológicas en un corpus extenso de lengua general

The automatic systems which deal with term’s extractions constitute an important tool when they make reference to the labor of compilation of lexemes, which is restricted to a specific field or specialty. The textual analysis that are realized for this type of software must include strategies that could detect collocations in the field in which is done. In this topic is studied the viability of the use from extensive textual’s corpus, that have not contain linguistic information, as happen with those textual’s corpus that could be compiled from internet. The internet is used like a source of information for the recompilation of terminology’s collocations. With that purpose is analyzed the behavior of different indicators based on the frequencies registered for a collection of economic terms in a Spanish corpus of 300.000 words.