Unigrams

Part of speech: noun

Definitions

1. Single words treated as separate units when analyzing or counting the pieces that make up a body of text
2. The individual words in a text, taken one at a time as the smallest units in statistical language analysis
3. One-word sequences used in computational linguistics to measure how often each separate word occurs

Etymology: "Unigrams" is the plural of "unigram," a term for a single unit in a sequence, most often a single word in a text being analysed. The word is built from two parts: the Latin-derived prefix "uni-," meaning "one," and "-gram," from Greek "gramma," meaning "letter" or "something written." Put together, the components describe a single written mark or item, and the sense has since broadened to any single element of a sequence. The "-gram" element is a familiar one in modern scientific vocabulary, appearing in words such as "telegram," "diagram," and "monogram." Its shared root gives "unigram" a family resemblance to these words, though the newer coinage belongs to the era of modern linguistics and text analysis rather than to ancient usage. Scholars building on the same pattern produced "bigram" and "trigram" for sequences of two and three units.