Topics
The distributional hypothesis is a linguistic theory that states that words that appear in similar contexts tend to have similar meanings:Â
This is based on the idea (by Firth in 1950s) that words are characterized by the company they keep. For example, words like “whale” and “dolphin” are often found near words like “ocean”, “sea”, “animal”, and “fish”. In distributional semantics, the similarity between two words is calculated by comparing the similarity between their vectors (e.g. dot product).Â