Why Vector Representations?

Computers natively understand 1s and 0s. Even the text displayed on computer screens is encoded in some numeric form. To make the processing easy, text is encoded as numbers. For the field of NLP, the demand is even more onerous. Here the computers are being taught to read, listen, and understand natural language. Mathematical functions are also being applied to the text data with the goal of detecting patterns.

NLP algorithms require large volumes of text data. However, the processing of this data takes a huge amount of time and eventually affects the performance of the algorithm. Thus, in order to make the processing faster and performance reasonable, we can take advantage of data structures. By representing data as vectors, we allow CPUs to operate over data in batches, which in turn improves performance. This is another key reason for representing text as vectors.

Encoding

The process of converting data into a specified format is called encoding...