An evaluation of MorphInd's morphological annotation scheme for Indonesian. Issue 2 (August 2021)

• Record Type: Journal Article
• Authors: , Prihantoro
• Is Part Of: Corpora. Volume 16:Issue 2(2021)
• Page Start: 287
• Page End: 299
• Publication Date: 2021-08
• Subjects: annotation -- Indonesian -- MorphInd -- morphology -- morphosyntactic; Corpora (Linguistics) -- Periodicals; 410.188
• DOI: 10.3366/cor.2021.0221
• Languages: English
• ISSNs: 1749-5032

Abstract

Abstract : MorphInd 2 (Larasati et al ., 2011 ) is a state-of-the-art morphological analyser for Indonesian. To date, there has not been any comprehensive evaluation of the morphological annotation scheme which MorphInd implements. My evaluation of this annotation scheme reveals a number of significant drawbacks. Some analytical features encoded in MorphInd's tagset seem not to reflect features actually present in Indonesian morphology, while certain common features in the analysis of Indonesian are absent. Likewise, the Part of Speech (pos ) hierarchy in the MorphInd tagset does not reflect the usual pos hierarchy used by Indonesian reference grammars. Moreover, the MorphInd output does not link morphological tags to the corresponding morpheme. Finally, a number of issues which might problematise text/corpus querying in the annotation's layout are observable, particularly relating to affixes, reduplication, and the affix–reduplication interface.