Minder Reader (中文: 读心机器人) is an artificial intelligence game which leveraging big data mining & crowd-sourcing to build intelligence bot.
The bot has been played by over 50 million times. The game itself have been deployed in multiple platforms, including iOS, Android, Windows Phone, Windows 10, and Web. The data mining technology has been transferred into many Microsoft products, such as Bing, Windows, and Cortana.
The bot is built on the knowledge mined from Renlifang (人立方), a relationship search engine in China which mine knowledge from billions of Web pages. Entities and relationships are automatically mined from the Web pages as the answers to Minder Reader.
The Minder Reader is devised to collect knowledge for various types of entities, such as people, movies, products or gifts, events, and food. In the game, the answerer chooses an entity but does not reveal it to the questioner, who will use as many as 20 yes/no questions to guess the correct answer. A user could play with the project’s mind-reader robot—the human-computer mode—or another human user—the human-human mode. In the latter, the bot will act as an assistant to provide hints when the questioner needs help. This project shows how knowledge is collected from the game and used in knowledge-mining algorithms. In addition, the potential of using the collected human-human question-answer data to train natural dialogue systems will be demonstrated.