![]() | Only 14 pages are availabe for public view |
Abstract This study examines the adequacy of the only two existing abstract meaning representation applications, namely The Universal Networking Language (UNL) and Abstract Meaning Representation (AMR) in their description of the meaning embedded in utterances or sentences. This is tested through comparing a semantically annotated corpus analyzed by both applications. The researcher of the present study employs the UNL to semantically annotate a novella, namely The Little Prince in its English version, which has been annotated by the other application and publically released earlier. Once the semantic annotation of the novella ends, another phase of the study commences. This phase entails comparing the adequacy of both systems through juxtaposing the representations against criteria of language features of the English language to avoid any bias to any of the two systems. The results of the comparison help arrive at a decision concerning the adequacy of both systems to realize the language features in their analysis and consequently produce reliable representations of meaning, especially that no other study has the same or similar aims. |