A Novel Automatic Definition Selector and Scoring Scheme for the Identification of Filipino Metaphors

Studies identifying metaphors are assessed per word as a scoring scheme for the effectiveness of the demonstrated methods. This scoring poses a problem in the validity of the interpretation of what a metaphor is. A metaphor is a figure of speech that relies heavily on the context and combination of words used to express an abstract concept. Dealing with metaphors embedded in the text can introduce inconsistencies in data processing since metaphors are known to change the meaning of the text. This challenge is even more prominent in high-context languages such as Filipino. This study presents a novel scoring scheme for assessing the identification of Filipino metaphors from social media data. In response to this issue, several more novelties were introduced in this study. These novelties are the Filipino metaphor corpus and the Filipino metaphor identifier created that utilizes an automatic definition selector.