automatic text generation (Natural Language Generation)
Automatic text generation, also known as Natural Language Generation (NLG), is a branch of artificial intelligence. It is part of Natural Language Processing (NLP) and refers to the generation of natural language text using software. Automatic text generation programs are increasingly used to create content for websites and other online applications. NLG software often uses a combination of machine learning and linguistic algorithms to generate text that follows a particular pattern or template. This requires structured data.
How NLG works
Natural Language Generation is a multi-stage process, with data being refined at each stage to generate natural language text. Automatic text generation consists of the following stages:
Filtering the data
Filtering the data is necessary to determine what should be included in the content generated at the end of the process.
Interpreting the data
The data is analysed and contextualised. This stage involves looking for possible patterns, often using machine learning.
Structuring the document
A document plan is designed and a narrative structure is selected based on the type of data to be interpreted.
Key sentences or parts of sentences are grouped together to summarise the topic.
Grammatical rules are applied to generate natural language text. The software is able to infer the syntactic structure of a sentence.
The final output is generated from a template selected by the user.
Applications of NLG
In principle, there are many areas where automatic text generation can be used to good effect. However, the speed of NLG software makes it particularly suitable for creating time-critical articles on the Internet (e.g. in news journalism).
NLG can also be used to create product descriptions or to generate intelligent responses for chatbots and voice assistants. NLG software also enables the transformation of complex business data into understandable content and the personalisation of responses to customer emails and messages.