{"id":37462,"date":"2022-01-26T09:00:00","date_gmt":"2022-01-26T07:00:00","guid":{"rendered":"https:\/\/forklog.com\/en\/?p=37462"},"modified":"2025-08-29T17:21:30","modified_gmt":"2025-08-29T14:21:30","slug":"what-is-natural-language-processing","status":"publish","type":"post","link":"https:\/\/forklog.com\/en\/what-is-natural-language-processing\/","title":{"rendered":"What is natural-language processing?"},"content":{"rendered":"
\n

What is natural-language processing?<\/strong><\/h2>\n

Natural-language processing (NLP) is a set of methods that help a computer system understand human speech.<\/p>\n

NLP is a branch of artificial intelligence<\/a>. It remains one of AI\u2019s hardest problems and is still far from fully solved.<\/p>\n<\/div>\n

\n

When did NLP emerge<\/strong>?<\/h2>\n

The roots of NLP go back to the 1950s, when the British scientist Alan Turing published the essay \u201cComputing Machinery and Intelligence\u201d<\/a>, proposing the \u201cTuring test\u201d<\/a>. One of its criteria is a machine\u2019s ability to interpret and generate human speech.<\/p>\n

On January 7th 1954, researchers at Georgetown University demonstrated machine translation. Engineers translated more than 60 Russian sentences into English fully automatically. The event spurred the field and went down in history as the Georgetown experiment<\/a>.<\/p>\n

In 1966, the American computer scientist of German origin Joseph Weizenbaum, at MIT, built the first chatbot, ELIZA. The program mimicked a conversation with a psychotherapist using active-listening techniques.<\/p>\n

\"What
A dialogue with \u201cEliza\u201d. Source: Wikipedia.<\/figcaption><\/figure>\n

In essence, the system paraphrased the user\u2019s messages to create the impression of understanding. In reality it did not grasp the substance of the dialogue. When it could not find an answer, it typically replied \u201cI see\u201d (\u201c\u041f\u043e\u043d\u044f\u0442\u043d\u043e\u201d) and steered the conversation elsewhere.<\/p>\n

That same year the Automatic Language Processing Advisory Committee (ALPAC) released a report<\/a> concluding that a decade of research had fallen short. Funding for machine translation was sharply cut.<\/p>\n

For the next decades breakthroughs were scarce, until the rise of the first machine-learning algorithms in the 1980s. Around the same time, statistical machine-translation systems appeared and research resumed.<\/p>\n

NLP boomed in the 2010s as deep-learning algorithms took off. Many of today\u2019s mainstays appeared then\u2014chatbots, autocorrect, voice assistants and the like. Recurrent neural networks became the default tool for many tasks.<\/p>\n

Another revolution came in 2019, when OpenAI unveiled the language model<\/a> Generative Pre-Trained Transformer 2, or GPT-2. Unlike earlier generators, it could produce long, coherent passages, answer questions, compose verse and even suggest new recipes.<\/p>\n

A year later OpenAI introduced GPT-3<\/a>, and big technology firms began showcasing their own large language models.<\/p>\n<\/div>\n

\n

How do NLP systems work?<\/strong><\/h2>\n

To answer that, consider how we humans use natural language.<\/p>\n

When we hear or read a phrase, several processes occur in parallel:<\/p>\n