Natural Language Processing (NLP) Concepts and Techniques

Vocabulary flashcards covering key terms, working concepts, components, and text normalisation preprocessing steps in Natural Language Processing.

Natural Language Processing (NLP)

An emerging technology and domain of AI that enables computers to understand, interpret, and generate human language.

Speech Recognition

The step in speech processing where a device listens to spoken voice commands and converts them into text.

Natural Language Understanding (NLU)

A component of NLP that helps machines understand and analyse human language by extracting metadata from the content, such as concepts, keywords, and emotions.

Response Generation

The process where an AI tool formulates a suitable response based on its understanding of the input request.

Speech Synthesis

The process of converting a generated text response back into spoken speech.

Knowledge Base

The source content with which words are compared to translate them and identify patterns or context in speech.

Natural Language Generation (NLG)

A component of NLP that acts as a translator converting computerised text into natural language to communicate back via speech or text.

Text Normalisation

A critical preprocessing step in NLP that involves converting text into a standardised format to clean and prepare textual data for analysis.

Sentence Segmentation

The first step in text normalisation that breaks a paragraph into different sentences to teach computers where one idea ends and another begins.

Tokenization

The process of breaking speech or text into smaller chunks called tokens, which can be individual words or clauses.

Tokens

Smaller chunks of speech or text, such as individual words, clauses, or punctuation marks, created through tokenization.

Stemming

A technique that simplifies a word into its root form by removing affixes, which can sometimes alter the actual word's meaning.

Affix

An addition made to the starting or ending of a word that determines its specific grammatical form.

Lemmatization

An accurate method that reduces a word to its base dictionary form (lemma) while keeping its context intact.

Lemma

The base or dictionary form of a word to which various inflected forms belong (e.g., 'sing' is the lemma for 'sings', 'singing', and 'sung').

Stop Word Removal

An NLP preprocessing step that removes common words with little or no relevance to the overall meaning of the speech or text.

Parts of Speech (POS) Tagging

The process following tokenization that assigns a part-of-speech category (such as Noun, Verb, Pronoun, Adjective, or Punctuation) to each token.