Word Map
The word map below shows the relationships among words with respect to their semantic and syntactic similarity. This visualization of words was developed using the two-layer neural network Word2vec function and T-distributed Stochastic Neighbor Embedding (t-SNE). Word2vec creates word vectors based on their semantic and syntactic meanings, and t-SNE represents these word vectors in two-dimensional space.
The model used to create the word map could also be used to find a list of words that are found frequently with a particular word. For example, the list of words below are often found with the word ‘steel.’ The probability that the word is associated with ‘steel’ is also shown. The word ‘jone’ represents the Jones and Laughlin Steel Company.
- jone, 0.6476
- corporation, 0.6083
- furnace, 0.5624
- mill, 0.5594
- produce, 0.5573
- work, 0.5291
- iron, 0.4713