In the last decade, analysis of parts of speech (POS) has reached a relatively stable stage while semantics still requires further studies and advanced techniques in the field of corpus linguistics. The field of semantics has focused more on the meaning of words. These phenomena are even more obvious in languages other than English. To differentiate from the present situation and facilitate the automatic parsing of thematic roles, this paper presents a developed parser based on the interaction among the thematic roles, grammatical functions, and parts of speech on three different linguistic levels, morphology, syntax, and semantics. The proposed syntax-semantic parser is created through computational programming, combining the design rules and the results obtained by other existing tools such as FreeLing and WordRrference, As a result this parser is capable of automatically analyzing different types of clauses and transforming the morphosyntatic information into the syntasemantic by importing data through users' interface. This study is limited to a few functions due to the difficulties of natural language processing and Spanish linguistics encountered during the collaboration of investigators from two research fields, information engineering and Spanish linguistics. Nevertheless, the results applied to the error detector and the suggestion system that were developed based on the learner' corpus the investigators established over several years, have academic value in the future.