AUTOMATIC SELECTION AND ANALYSIS OF VERB AND ADJECTIVE SYNONYMS FROM JAPANESE SENTENCES USING MACHINE LEARNING

In this study, we investigate a method to select appropriate verb and adjective synonyms automatically for a given sentence from multiple options based on machine learning. Our experimental results demonstrate that the accuracy rates of the proposed machine learning method for selecting verb and adjective synonyms (i.e., 0.84 and 0.76, respectively) are higher than those of a baseline method that always outputs the most frequently used synonyms from the training dataset (i.e., 0.70 and 0.67, respectively). Furthermore, the machine learning performance classifies the verb and adjective synonym pairs based on whether either of the synonyms could be used in a given sentence. Additionally, we examined important machine learning features for the selected verb and adjective synonym pairs and revealed useful information for selecting accurate synonyms. The performance of our machine learning approach was then compared with that of human test subjects. Finally, we discussed the differences among nouns, adverbs, and verbs/adjectives during synonym selection.