In linguistics and computer science, a sequence of two consecutive words or characters used in text analysis and language processing.
From the prefix 'bi-' (Latin, meaning 'two') plus 'gram' (Greek, meaning 'letter' or 'written character'); the term was coined in mid-20th century computational linguistics.
Bigrams are the secret weapon of spam filters, autocorrect, and search engines—they work because language isn't random; certain two-word pairs ('the quick,' 'New York') appear together far more often than statistics alone would predict.
Complete word intelligence in one call. Free tier — 50 lookups/day.