tokenisation

Breaking into units

Tokenisation is the process of breaking text into smaller units called tokens—typically words, subwords, or characters.

In natural language processing, tokenisation is a fundamental step before any text can be analysed or processed by algorithms: The model's tokenisation splits "unhappiness" into three pieces.

In finance and security, tokenisation replaces sensitive data with non-sensitive placeholders: Credit card tokenisation protects customer information.

Also spelled tokenization.

nlp

tokens

Currency substitutes

segmentation · subword · vocabulary · embedding · encoder · decoder · transformer · BERT · GPT · corpus

linguistics

morphemes

Smallest meaningful language units

morphology · inflection · derivation · syntax · semantics · segmentation · orthography · syllables · compounds · affixes

security

encryption

Converting data into code

authentication · digital payments · secure storage · credentials · obfuscation · data protection

programming

lexical analysis

Character-to-token conversion process

interpreter · tokens · regular expressions · abstract syntax tree · source code · keywords · identifiers · literals · operators · delimiters

Opposites
segmentation vs. combination

tokenisationconcatenation

Joining items end-to-end

joining · merging · aggregation · synthesis · combination

financial/security transformation

tokenisationdecryption

Converting ciphertext to plaintext

Family

token, tokenization, tokenizer

Familiarity
Top 1,000,000
Profile
Timeless · Abstract · Rated G