Creative Technology in personal practice is an opportunity to explore a piece of technology while learning it.
Based on my interest on natural and computational language, I choose to work with natural language tool kit (NLTK).
Specifically, I build a tool that enables people to get a a color palette out of a piece of text. To achieve this, I used the word2vec algorithm, which is typically used to learn word associations and linguistic context from a large corpus of text. Word2vec takes as its input a large corpus of text and produces a vector space, typically of several hundred dimensions, with each unique word in the corpus being assigned a corresponding vector in the space, so that a spatial relation is created between each word to represent their semantic relationships.
Specifically, I build a tool that enables people to get a a color palette out of a piece of text. To achieve this, I used the word2vec algorithm, which is typically used to learn word associations and linguistic context from a large corpus of text. Word2vec takes as its input a large corpus of text and produces a vector space, typically of several hundred dimensions, with each unique word in the corpus being assigned a corresponding vector in the space, so that a spatial relation is created between each word to represent their semantic relationships.
After learning about that, I decided to try visualizing this spatial connection using colors, since every color on screen has a RGB value with three coordinates of red, green, and blue, which means every color can be represented by a vector in a 3D dimensional space, so that I’m able to map every word into color. And I thought it would be more interesting to construct a personalize color zone using analogical words to shows the semantic relationships, instead of using name of a color.
