AI Sketchbook Series #3- Text Representation - Word2Vec

Here is the third installment of the AI Sketchbook Series, where we demystify fundamental AI concepts with a semi-visual approach. In our previous posts, we explored how to represent text using One-Hot Encoding and the Bag-of-Words TF-IDF method. While these techniques are a great starting point, they treat each word as a standalone entity, ignoring the rich context and relationships between words.

In this blog post, we'll dive into Word2Vec, a powerful and popular technique that represents words as dense vectors, or "embeddings." By doing so, we can capture the semantic and syntactic relationships between words, allowing AI models to understand context and meaning in a way that goes far beyond simple word counts.