Speech-enabled card games for incidental vocabulary acquisition in a foreign language

  • Authors:
  • Ian McGraw;Brandon Yoshimoto;Stephanie Seneff

  • Affiliations:
  • MIT Computer Science and Artificial Intelligence Laboratory, 32 Vassar Street, Cambridge, MA 02139, USA;MIT Computer Science and Artificial Intelligence Laboratory, 32 Vassar Street, Cambridge, MA 02139, USA;MIT Computer Science and Artificial Intelligence Laboratory, 32 Vassar Street, Cambridge, MA 02139, USA

  • Venue:
  • Speech Communication
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we present a novel application for speech technology to aid students with vocabulary acquisition in a foreign language through interactive card games. We describe a generic platform for card game development and then introduce a particular prototype card game called Word War, designed for learning Mandarin Chinese. We assess the feasibility of deploying Word War via the Internet by conducting our first user study remotely and evaluating the performance of the speech recognition component. It was found that the three central concepts in our system were recognized with an error rate of 16.02%. We then turn to assessing the effects of the Word War game on vocabulary retention in a controlled environment. To this end, we performed a user study using two variants of the Word War game: a speaking mode, in which the user issues spoken commands to manipulate the game cards, and a listening mode, in which the computer gives spoken directions that the students must follow by manipulating the cards manually with the mouse. These two modes of learning were compared against a more traditional computer assisted vocabulary learning system: an on-line flash cards program. To assess long-term learning gains as a function of time-on-task, we had the students interact with each system twice over a period of three weeks. We found that all three systems were competitive in terms of the vocabulary words learned as measured by pre-tests and post-tests, with less than a 5% difference among the systems' average overall learning gains. We also conducted surveys, which indicated that the students enjoyed the speaking mode of Word War more than the other two systems.