Named Entity Hunt Game

How to play: choose a label, drag across canvas words, then submit. Exact boundaries earn the highest score.

Game configuration

Game mode

Difficulty

Passage topic

Rounds

Time limit (seconds)

Starting lives

Hint allowance

Random seed

Player one

Player two

Boundary matching

Date format

Currency focus

Measurement system

Allow nested entities

Include punctuation

Case-sensitive matching

Instant selection feedback

Enabled entity categories

Custom passage and annotations

Paste text, then enter entities as JSON. Each entity needs text, type, and optional explanation.

Import TXT, CSV, or JSON

Entity hunting canvas

Configure the game, then press Start new game.

Ready

Keyboard: 1–9 labels, Ctrl+Z undo, Ctrl+Y redo, Enter submit, H hint.

Feedback and answer review

Correct selections, boundary mistakes, missed entities, explanations, and model comparisons will appear here.

Live scoreboard

Score

Time

00:00

Round

0/0

Lives

Streak

Hints

Current player

—

Solo game

Game progress0%

0%
Precision

0%
Recall

0%
F1 score

0%
Accuracy

Display and access

Dark mode

High contrast

Large text

Reduced motion

Sound effects

Achievements

Performance analytics

Round history and exports

Round	Player	Topic	Mode	Correct	Missed	Wrong	Precision	Recall	F1	Time	Score
No completed rounds yet.

Learning guide

Named entity recognition

NER locates meaningful spans and assigns labels. Strong annotations use consistent boundaries, categories, and formatting rules.

Precision and recall

Precision rewards clean predictions. Recall rewards finding every relevant entity. F1 balances both measures.

Common mistakes

Watch partial names, organisation-location ambiguity, date formats, titles, acronyms, products, quantities, and monetary expressions.