Semantris is a word association game powered by machine learning.

Semantic Experiences lets you get hands-on with games and experiments that showcase advances in natural language understanding.

Semantris is a word association game that uses this same technology. Each time you enter a clue, the AI looks at all the words in play and chooses the ones it thinks are most related.

In a traditional analog mad lib, one person has the template, another person supplies the words, and hilarity ensues. We can use the input() & print() functions, along with concatenation, to implement a …

The orientation predictors can detect the overall orientation of a document image or word crop. They return the general orientation — [0, 90, 180, -90 (270)]— along with the corresponding confidence …

Essentially, we will be playing a big game of "guess the missing word". For each input sample we will obscure 25% of our input data, and train our model to predict the parts we covered up.

We will implement Word-level tokenisation, where each word in the text corpus will be translated into its own token. This approach has its drawbacks; it struggles with rare words and requires a large …

An article of 1000 words can be summarized to both 200 words and 100 words depending on its content. In 2014, Cho et al. and Sutskever et al. proposed to use an encoder-decoder model purely based on …

Blocks of text and blocks of code. In order too create a block of text or code just hover over the cells and click on the words + Text or + Code above or below the cell. Also you may edit and see the text/code …

To run the code only in this cell, you can use the keyboard Shift-Enter when you've clicked into it or hit the play button on the left side of the cell. Once this code is run, you'll see a number in the top left …