NooJ Dictionary for Rromani: Importing of an Editorial Dictionary to the NooJ System

This article presents the process of importing an editorial dictionary [1] into the NooJ system. The current module for Rromani contains only 640 dictionary entries. However, we have written the central inflectional grammars of nouns, verbs, adjectives and grammatical words using the NooJ environment. Now, we import the editorial dictionary to complete the NooJ dictionary. Rromani language has a complex dialect structure, yet its dialectal equivalents show systematic correspondences between them [2]. This fact encourages us to develop a single and joint module for the entire Rromani language. The editorial dictionary mentioned above includes variants of the four Rromani dialects. The problem is that the information related to part of speech (POS) categories is absent or not precise enough to be processed by the NooJ parser. We have made implicit information explicit and identified 2,559 nouns, 830 adjectives and 686 verbs among 4,524 entries. However, we still have to identify the categories of other 448 words by referring to additional dictionaries [3–5]. We would first complete importing the entire editorial dictionary and then examine each entry to perfect the inflectional morphology and formalize the derivational morphology in the NooJ module for Rromani.