akoksal/Turkish-Word2Vec

Trained on Turkish Wikipedia corpus using gensim, the model handles agglutinative morphology by learning vector representations across inflected word forms and suffixes. Includes downloadable pre-trained embeddings and demonstrates semantic operations like analogy tasks (king+woman-man=queen) adapted for Turkish grammatical structures. A lemmatization module is planned to improve quality by consolidating morphological variants.

222 stars. No commits in the last 6 months.