What is NLG (Natural Language Generation) in 3 minutes
Published April 25, 2019·Written by Wonderflow
You might be familiar with NLP, but do you know what is NLG? In today’s video, our CEO Riccardo Osti will explain the meaning of Natural Language Generation, and its relation with NLP. NLG and NLP are closely related, since Speech Recognition is a subfield of NLP, or to be more precise, it is a subfield of computational linguistics. So if you are interested in this topic, in AI and/or machine learning, watch it right now!
Today I will explain what NLG, natural language generation is in a couple of minutes. First of all, I am speaking about this right after NLP because Speech Recognition is a subfield of NLP. Or, to be more precise, it is a subfield of computational linguistics.
The goal of Natural language generation (NLG) is to use AI to produce written or spoken narrative from a dataset. NLG is what enables machines and humans to communicate seamlessly, simulating human to human conversations. In other words, NLG uses numerical information and mathematical formulas to extract patterns from any given database and translates them into a text that is easy for humans to understand.
A good example of NLG is automated journalism. Where a computer searches the web for real-time news, scapes the data from different sources and writes a text summary, that can be published very quickly to the web.
NLP and NLG
As we said, NLG is related to NLP and natural language understanding. In many cases, these two activities are connected to each other. For example, when we interact with a chatbot we usually start asking a question, and the computer decides which answer to give. This happens based on the input that we provided.
In this process, the most difficult part is in the first step. At this step, the machine needs to understand what we want, where the easiest part is surprisingly natural language generation, where the machine creates the textual answer. In fact, even if NLG seems more spectacular and futuristic compared to text analysis, it is, in reality, easier than what we think. How is this possible?
In text analysis, the machine has to deal and interpret all kinds of unknown, ambiguous or erroneous expressions. While in NLG, the ideas that the system transforms into texts are generally known precisely.