Word Embeddings - From PPMI to Contextual Representations - L3

Learn how words are represented as high-dimensional vectors in this comprehensive lecture covering the evolution from PPMI representations through word2Vec and GloVe to modern contextual embeddings like BERT. Explore the fundamental concepts of word embeddings and understand how different approaches to defining context shape representational hypotheses in natural language processing. Examine the progression of embedding techniques and discover how contextual information is encoded to create meaningful vector representations of words that capture semantic relationships and linguistic patterns.