Evaluation of corpus-assisted Spanish learning

In the development of corpus linguistics, the creation of corpora has had a critical role in corpus-based studies. The majority of created corpora have been associated with English and native languages, while other languages and types of corpora have received relatively less attention. Because an increasing number of corpora have been constructed, and each corpus is constructed for a definite purpose, this study identifies the functions of corpora and combines the values of various types of corpora for auto-learning based on the existing corpora. Specifically, the following three corpora are adopted: (a) the Corpus of Spanish; (b) the Corpus of Taiwanese Learners of Spanish; and (c) the Parallel Corpus of Spanish, English, and Chinese. These corpora represent a type of native, learner, and parallel language, respectively. We apply these corpora as auxiliary resources to identify the advantages of applying various types of corpora in language learning from a learner’s perspective. In the environment of auto-learning, 28 participants completed frequency questions related to semantic and lexical aspects. After analyzing the questionnaire data, we obtained the following findings: (a) the native corpus requires a more advanced level of Spanish proficiency to manage ampler and deeper context; (b) the learners’ corpus facilitates the distinction between error and correction during the learning process; (c) the parallel corpus assists learners in connecting form and meaning; (d) learning is more efficient if the learner can capitalizes on specific functions provided by various corpora in the application order of parallel, learner and native corpora.