Artificial Intelligence is no more just a buzzword, covid pandemic accelerated the adoption of digitization & Artificial intelligence in different sectors of life, which raises the new set of problems of "IT jargons" which perplex & excites at the same time. One of the most common uses terminologies in the conversational domain, which confuse everyone is NLP & NLU, sometimes used interchangeably without understanding the difference.
Recently, I did some research on this topic in preparation for a client demo, where we were presenting our Chabot solution. Understanding these terms is essential for the development of any conversational applications, e.g. Chabot and IVR bots.
What is NLP?
Natural Language Processing (NLP) is a core part of conversational AI where a large volume of data is processed & broken down in smaller elements to understand the relationship between the smaller sets, obviously as the name suggests it is related to language processing. In order words, the NLP process & turns text into structure data.
What is NLU?
Natural Language Understanding, as the name suggests, enables the computer to understands the meanings of the input (Text, Speech) given to it. NLU uses AI algorithm to process natural language to understand the sentiment, intent, semantics & context of the given input.
In Conversational applications, NLP plays a critical role, while, NLU is the subset of NLP.
For Example
During the conversation, if you greet someone, there are different ways of greeting people according to their culture and norms. If I say hello, Hallo, Good morning or Salam, Shalom, Namaste which conveys the same meaning, i.e. "greetings" but communicated in different ways, using NLU, the computer will recognise the intent of the conversation which is in this case "greetings".
Similarly, if we ask someone "How are you?" there are different ways of asking it.
- How's everything?
- How's it going?
- How are things?
- What's up?
- How are you doing?
Here again, NLU plays its part in understanding the intent, even if the user inputs the wrong spellings, write fragmented sentences or pronounce wrong words, NLU will be able to process it. Whereas, NLP analyse text and speech, focusing primarily on language structure. However, it will not tell you what was meant or intended by specific input.
NLP looks what was said, whereas, NLU looks at what was meant.
I hope now you understand the difference between NLU and NLP. Let me know in comments what your opinion is?
Comments