Character-Level Morphological Analysis in Turkish

This is a proposed model for Neural Morphological Analysis. The model takes input character embeddings of a token and outputs lemma, part-of-speech, and set of morphological features.

TODO:

• Contextual Embeddings
• Finite State Transducer for Phonological Normalization
• Reducer Model implementation
• Dataset Comparison
• Model Comparison