
Submission + - Computer learns language by playing games (mit.edu) 1
Frans Faase writes: "By basing its strategies on the text of a manual, a computer infers the meanings of words without human supervision. The paper Learning toWin by Reading Manuals in a Monte-Carlo Framework explains how a computer program succeeds in playing Civilization II using the official game manual as a strategy guide. This manual uses a large vocabulary of 3638 words, and is composed of 2083 sentences, each on average 16.9 words long. By this the program improves it succes rate from 45% to 78% in playing the game. No prior knowledge of the language is used."