NLP Explained In Five Minutes
The Foundation
As you might be able to tell by now, I'm interested in where data analytics and marketing intersect. But the reality is that these topics are pretty complicated. Sentiment analysis, intention detection, machine learning, text analytics―all of these phrases can begin to sound like plot devices in a Douglas Adams novel. But understanding them is vital if you want to bring your business and marketing apparatus to the cutting edge of the 21st century. Which brings us to the source from which all these tools flow: Natural Language Processing (NLP).
What is NLP?
At its most fundamental Natural Language Processing is interacting with computers like we interact with other humans. Without NLP most of data science couldn't exist. Natural Language Processing is the study of making computers understand how humans naturally speak, write and communicate. Armed with this understanding, these computers are able perform various analyses on a huge scale, providing professionals with meaningful insights.
Traditionally, communicating with a computer would require giving it very precise, unambiguous, and highly structured instructions, written in dedicated programing languages, like Java, C++ and Ruby.
Java? Whatevs...
But us humans don’t speak to each other like we speak to computers.
We don't always follow the rules. Human communication conveys messages in ways that, while structured with grammar, can be imprecise and ambiguous. Often, like with slang or idioms, words and their meaning can vary region to region in the same country. NLP aims to bridge the gap between these vast differences in communication.
Learning Like a Machine
Modern techniques and approaches for NLP are based on what is called machine learning. Machine learning in NLP happens when we ask an artificial intelligence to examine patterns within data to draw conclusions on how natural human languages work. By applying these conclusions, machines are able to perform tasks including sentiment analysis, text parsing, speech recognition, part-of-speech tagging, and more.
Consider the following sentence for a second
With current NLP methods, machines will break down this tweet into its grammatical elements (“amazing” = adjective, “Cloud” = noun, “delivers” = verb, etc.). In this case Cloud is referencing Cloud Computing. Of course, ASAP is the common acronym for As Soon As Possible. Using this information, Natural Language Processing provides the foundation for further text analytics, like intention detection, sentiment analysis and other linguistic analyses.