Computational Linguistics Explained

This image was created with the assistance of DALL·E 3

When you read the phrase “computational linguistics,” the first thing that comes to your mind will probably be MIT, Harvard or Ivy League graduates sitting in a room in front of computers finding every single piece of information regarding a language or group of languages. However, computational linguistics is more prevalent in your daily life than you might first come to expect. Anything from asking a voice assistant to turn the thermostat on or to translate a phrase from one language to another on the fly involves computational linguistics.

What is Computational Linguistics?

The textbook definition of computational linguistics would be:

The scientific and engineering discipline concerned with understanding written and spoken language from a computational perspective, and building artifacts that usefully process and produce language, either in bulk or in a dialogue setting. (Schubert 2014)

However, this definition is too word-heavy and abstract for most people to understand. So let's break it down phrase-by-phrase in order to gain an understanding of the entire definition

“The scientific and engineering discipline concerned with understanding written and spoken language from a computational perspective” (Schubert 2014)

This phrase can be roughly interpreted as “The field of understanding language (in any way, shape or form) like a computer.” When you delve even further into this definition, it means understanding language not as a biological human but as a machine which only thinks logically. Now, lets interpret the next part of the quote

“ ..and building artifacts that usefully process and produce language, either in bulk or in a dialogue setting. (Schubert 2014)

This phrase can be interpreted as “Building apps that process and write stuff in a language as a chatbot or as in a translation service.” Again, when you analyze this definition further, it boils down to creating apps that can manipulate, combine or create information in the form of a language.

When you put the two analyzes together, a definition of computational linguistics in a modern and simple context is produce. That definition is, “The field of understanding language (in any way, shape or form) and creating apps that can manipulate, combine or create information in the form of a language.”

Why is Computational Linguistics Necessary?

Computational linguistics is an important aspect of our modern technological lives. From Google Translate to Autocorrect, below are some applications for computational linguistics.

• Translating: When you open up an app like Google Translate and Translate.com, on your phone, use computational linguistics to translate words and phrases from one language to another while keeping grammatical structure in mind.
• Retrieving Information: Even when you use Google or DuckDuckGo or any other search engine, computational linguistics helps split up and understand your query and bring you relevant results.
• Recognizing Your Words: When you ask Siri, Alexa, or Google Assistant for help, computational linguistics is at work, understanding your spoken words into a format the computer will understand and do something with.
• Predicting What You're Going To Say: When you're going to type a message to your best friend, computational linguistics corrects almost all of your misspelled phrases, learning from the way you and the people you talk to use language.

How does Computational Linguistics Work?

Now that you know what is computational linguistics and why it's necessary in the 21st century, let's take the next step in the logical route of progression and talk about how it works. If necessary and if enough people comment so, I may write a blog post detailing its implementation with Keras and TensorFlow.

Rambling aside, below are some of the tools that computational linguists (people who work in the field of computational linguistics) use in their average day at work:

• Corpora: Large, structured sets of text in a particular language (preferably with translations in an already known language) which act as the training grounds for computational linguistics models. They act as textbooks in the metaphorical classrooms of computationally linguistic models. (Li 23)
• Machine Learning: This subset of artificial intelligence is all about learning from data. By exposing a computer to enough examples, it can start to recognize patterns and make predictions. (Coursera 23)
• Natural Language Processing (NLP): This is the technology that converts human language into a form that computers can understand, be it written or spoken. (IBM)
• Syntax Trees and Parsing: These are methods to break down sentences into their grammatical components, helping computers to understand the structure and meaning of sentences. (McRoy 21)

What this Post Means for the Future

Computational linguistics helps computers understand human language, despite its changes and quirks. The goal of computational linguistics is to make it easier to talk with machines. It’s making it easier for us to chat with machines, blending language and tech in helpful ways. I hope you learn the impact of computational linguistics in our daily lives and come to appreciate just how complex asking a voice assistant to close your garage door.

Citations (MLA):

Schubert, Lenhart. “Computational Linguistics.” Stanford Encyclopedia of Philosophy, Stanford University, 26 Feb. 2014, plato.stanford.edu/entries/computational-linguistics/.

Li , Michelle. “Research Guides: Linguistics: Linguistic Corpora.” Linguistic Corpora — Linguistics — Research Guides at UCLA Library, University of California, Los Angeles , 23 Oct. 2023, guides.library.ucla.edu/c.php?g=180293p=1189870#:~:text=A%20computer%20corpus%20is%20a,David%20Crystal.

Coursera. “Machine Learning vs. AI: Differences, Uses, and Benefits.” Coursera, Coursera Inc, 2 Nov. 2023, www.coursera.org/articles/machine-learning-vs ai#:~:text=Machine%20learning%20,tasks%20without%20any%20human%20intervention.

IBM. “What Is Natural Language Processing?” IBM, www.ibm.com/topics/natural-language-processing#:~:text=Natural%20language%20processing%20,machine%20learning%2C%20and%20deep. Accessed 7 Nov. 2023.

McRoy, Susan. “Data Structures and Processing Paradigms.” Principles of Natural Language Processing, Susan McRoy, 24 July 2021, uwm.pressbooks.pub/naturallanguage/chapter/chapter-2/#:~:text=The%20data%20structures%20most%20common,be%20discussed%20in%20Chapter%204.