Text to Emoji Converter

Emoji communication carries measurable information density. A single pictograph can replace 1 - 3 words while preserving semantic intent, but naive word-to-emoji substitution produces incoherent output. This converter applies greedy longest-match tokenization across a dictionary of 800+ mapped terms, with basic stemming (suffix stripping of -ing, -ed, -ly, -s) to catch inflected forms. Multi-word phrases like “ice cream” resolve to a single emoji before individual words are evaluated. The conversion ratio matchedtotal quantifies coverage. Three output modes let you control density: hybrid text, pure emoji, or annotated with original-word tooltips.

Limitations exist. The dictionary covers common English vocabulary. Proper nouns, slang, and domain jargon will pass through unconverted. Homonyms resolve to the most frequent sense only (“bat” → 🦇, not a cricket bat). Context-dependent meaning is not analyzed. Pro tip: shorter, concrete sentences convert better than abstract prose.

The converter applies a greedy tokenization algorithm. Input text is scanned left-to-right. At each position, the longest matching phrase in the dictionary is consumed first.

Where D is the emoji dictionary (hash map), w is the current token (lowercased), and stem strips common English suffixes: -ing, -ed, -ly, -er, -est, -s, -es, -tion, -ment, -ness.

The conversion ratio is computed as:

Multi-word phrase detection uses a sliding window of up to 4 tokens. The algorithm checks 4-word, then 3-word, then 2-word combinations before falling back to single-word lookup. This ensures “ice cream” maps to 🍦 rather than 🧊 + unconverted “cream”.