# text (110-text.com) Text is the words people read and write. It is messages, books, emails, and anything made of letters and sentences. ## What a model may hear - plain string data (programming and databases): treat it as raw characters without formatting, not a number or date - SMS message (telecommunications): assume a short 160-character message sent to a phone number - machine-readable annotation (machine learning datasets): expect labeled training examples paired with categories or answers - unstructured document (information retrieval): prepare to index, search, or extract from large bodies of words - rendered glyph sequence (computer graphics and fonts): consider font files, layout engines, and display properties ## Where people and models part ways - Says: "Send me the text" Means: Email or message me the information May be taken as: Format it as a raw code string or SMS protocol payload Say instead: "Email me the information in plain words" - Says: "Analyze this text" Means: Tell me what this article says May be taken as: Run statistical NLP classification or sentiment scoring without summary Say instead: "Summarize what this article means in plain language" - Says: "Make it text-only" Means: No images or videos, just words May be taken as: Strip all markup and structure including headings and paragraphs Say instead: "Use only words, but keep headings and paragraphs for readability" - Says: "Extract the text" Means: Pull out the readable words from this file May be taken as: Dump raw OCR output or binary-encoded characters without cleaning Say instead: "Pull out the readable words and fix any obvious scanning errors" ## Tips - Say 'plain words' or 'sentences' if you do not want technical string handling - Say 'summarize' or 'explain' instead of 'analyze' unless you want statistics - Mention 'headings' and 'paragraphs' if you want structure kept - Say 'message' or 'email' instead of 'text' if you do not mean SMS - Say 'readable words' if you want human-oriented output, not code ## Often confused with - string: a programming data type, not natural language - content: broader, includes images and video - copy: specific to marketing or published writing - script: executable code, not just readable words - document: implies formatting and file structure - corpus: a collected body for linguistic study