Distributional similarity holds that words appearing in similar contexts tend to have related meanings. Embeddings turn this principle into vectors: words sharing contexts receive nearby representations, supporting semantic search. section 7.2.2
Context recovers meaning
In “The customer finally signed the cotnratc,” the surrounding words suggest “contract.” Similarly, apple and banana appear in contexts such as fruit, juice, and healthy, so their embeddings are closer than car’s. section 7.2.2 section 7.2.2
Figure 7.3 Semantically similar words are closer to each other in embedding space.
Use the embedding-space diagram to see how semantically similar words are positioned near one another. section 7.2.2
