https://towardsdatascience.com/visualization-of-word-embedding-vectors-using-gensim-and-pca-8f592a5d3354

T-SNE

https://habr.com/en/company/vk/blog/449984/

https://github.com/sismetanin/word2vec-tsne/blob/master/Visualizing Word2Vec Word Embeddings using t-SNE.ipynb

https://www.kaggle.com/mateiionita/visualizing-bert-embeddings-with-t-sne

코랩 구현

https://colab.research.google.com/drive/1xv99wTySfsJPExHNzl1s5g9DbQc0RKhl?usp=sharing

→ 각 단어마다 top 30개 similarity 뽑아서 시각화함

문제점) 단어 갯수 몇개 안되는데도 램 엄청 잡아먹음 → 전체 뉴스 데이터 비효율적

K-means

https://dylancastillo.co/nlp-snippets-cluster-documents-using-word2vec/

문제점