adamsamson/HornMorpho2.5

HORNMORPHO is a Python program that analyzes Amharic, Oromo, and Tigrinya words into their constituent morphemes (meaningful parts) and generates words, given a root or stem and a representation of the word’s grammatical structure. # HornMorpho 2.5 HORNMORPHO is a Python program that analyzes Amharic, Oromo, and Tigrinya words into their constituent morphemes (meaningful parts) and generates words, given a root or stem and a representation of the word’s grammatical structure. It is part of the L3 project at Indiana University , which is dedicated to developing computational tools for under-resourced languages. Natural language applications, such as question-answering, speech recognition, information retrieval, and machine translation, rely on a lexicon of possible forms in the relevant language. Morphological analysis is important for morphologically complex languages like Amharic, Oromo, and Tigrinya because it is practically impossible to store all possible words in a lexicon, and many words have close to 0 probability of occurrence in any given corpus. This becomes obvious in the context of machine translation to a morphologically simple language such as English, where the correspondence between words in Amharic, Oromo, or Tigrinya and the other language will often be many-to-one. The Amharic word ባይከፈትላቸውም, for example (which incidentally does occur in an online corpus), could be translated as ‘even if it isn’t opened for them’. While a system for processing English could include all of the English words in the translation (even, if, it, isn’t, opened, for, them) in its lexicon, an Amharic system that includes all words such as ባይከፈትላቸውም is clearly impractical. For translation into Amharic, Oromo, or Tigrinya and sophisticated question-answering, morphological generation is also desirable because it is probably impossible to store all of the words that the system will output. Finally, for Semitic languages such as Amharic and Tigrinya, morphological analysis can make explicit some of the phonological features of the languages that are not reflected in the orthography; these features may be important for text-to-speech applications. For example, the Tigrinya word ዚፍለጥ ‘which (it) is know …