Climbing towards NLU (Natural Language Understanding)
Don’t be confused! The title is right. Not NLP, but NLU for “Natural Language Understanding”. NLU is a subfield of NLP and helps computers understand and interpret human language.
More specifically, NLU is defined by Gartner as “the comprehension by computers of the structure and meaning of human language (e.g., English, Spanish, Japanese), allowing users to interact with the computer using natural sentences”.
The success of the large neural language models on many NLP tasks is exciting. However, we find that these successes sometimes lead to hype in which these models are being described as “understanding” language or capturing “meaning” — Bender, Emily M., and Alexander Koller. “Climbing towards NLU: On meaning, form, and understanding in the age of data.” Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020.
NLU & NLP: What is the difference?
The definitions of NLP and NLU are many times confused. NLU is one of the components of natural language processing. More specifically, it is a subset of understanding the part of NLP.
NLP tries to analyze and understand the text of a specific document, while NLU tries to develop a dialog between human and computer machines.
Both, NLU and NLP, understand the human language. However, NLU communicates with untrained individuals to learn to understand which is the intention of the writer or speaker. One of the challenges of NLP is the human errors, like misspelling or transposed letters and words, in writing. NLU tries to understand words and meaning of documents besides these human errors.
In simple words, NLP looks at “what” was said, while NLU looks at what was “meant”.
NLU Applications
Do you know which applications are based on Natural Language Understanding? Let’s take a look in some of them.
Voicebot
Chatbots and voicebots like Apple’s Siri and Amazon’s Alexa understand the human language. These applications use both NLP and NLU, in order to take the desired results.
Sentiment Analysis
Ironic and sarcasm is very common while someone chatting. Many times it is not clear if a word has a positive or negative meaning. For example the phrases “or right” or “whatever” can have both positive and negative meaning, depending on the intent of the writer or speaker.
These days Sentiment Analysis is used especially in Sales and Marketing in order to understand customer reviews.
Question Answering
Nowadays, most of the devices give the opportunity to communicate with users (humans), using the human language. Regardless of the language you speak your Google Assistant can be your savior.
Voicebots can question answering by NLU, while Google Assistant can interpret 44 languages and it can process both verbal and written queries.
Sources:
Take a look here and follow me on social media: https://linktr.ee/codingggirl
