Understanding Word2Vec: Teaching Neural Networks to Understand Words

As humans, we use words to communicate, but in the case of machines we are usually accustomed to hearing about 0s and 1s . Now what about machine learning algorithms , which are essentially trying to simulate aspects of how our brains work? Many machine learning algorithms, including neural networks , do not work well with raw words . So if you want to plug words into a neural network or some other machine learning algorithm, we need a way to convert those words into numbers . One simple way to convert words to numbers is to assign each word a random number . For example, someone watched the movie The Incredibles and said: "The Incredibles is great!" We can give each word a random number. Now, if we want to create another sentence like: "The Incredibles is awesome!" We can reuse the existing numbers and only assign a new number for "awesome" . However, this approach has a disadvantage. Even though "great" and "awesome" have very similar meanings, the numbers assigned to them are comple