Practical AI: Using NLP word vectors in a novel way to solve the problem of localization

The most practical use of word embeddings (word2vec, glove, etc) you will ever see.

Image by Author

King — Man + Woman = Queen

You might have seen the traditional word2vec or Glove word embeddings examples that show King -Man+Woman = Queen. Here Queen will be returned from the word embedding algorithm given the words King, Man, and Woman. Today we will see how we can use this structure to solve a real-world problem.

1. The problem definition:

An edtech company in the USA wants to expand into India after being successful in its home market. It has a large set of questions in their question bank that it wants to use when it enters the Indian market.

But there is one big problem. A sample third class (grade) math question in their question bank looks like this —

Frank lives in San Francisco and Elizabeth lives in Los Angeles. If the flight time is 2 hrs when will Elizabeth reach Frank if she starts at 8am in the morning?

A 3rd-grade kid living in India would not connect with this question as it has references to names and locations lesser know to him/her - Frank, San…