x-lai / analogy-finisher-and-word-similarity-calculator---natural-language-processing Goto Github PK
View Code? Open in Web Editor NEWThis project forked from mohammadhia/analogy-finisher-and-word-similarity-calculator---natural-language-processing
A project that develops an analogy finisher (predictor of final word in an analogy) by using various word embeddings as training (such as the GloVe and LexVex embeddings; which are not included here due to the size of the files). Mikolov's analogy dataset was used to test the model's accuracy in completing the analogies. Additionally, the most similar words (synonyms) and most dissimilar words (antonyms) were calculated using cosine similarity for a select number of words.