Project LISTEN

Project LISTEN (Literacy Innovation that Speech Technology ENables) at Carnegie Mellon University. It has been supported by National Science Foundation under IERI^[1] and ITR^[2] programs, but currently it is supported by the U.S. Department of Education’s Institute of Educational Sciences^[3] under Grants R305B070458, R305A080157 and R305A080628, and by the Heinz Endowments.^[4] The Project is an initiative to create a novel tool to improve literacy. It works like an automated Reading Tutor which displays stories on computer screen and simultaneously listens to children read loud. This project is described as ‘an automated reading Tutor’.^[5] The project is headed by David 'Jack' Mostow, Ph.D. It has used the cepstral N-gram stochastic and language-level models of CMU Sphinx to evaluate oral reading and speaking proficiency and provide literacy tutelage.

This project contributes to and provides unique opportunities for educational data mining. It is not a commercial project yet, but has been used by hundreds of children as part of its research and testing.^[6]

In 2005, the LISTEN reading tutor was pilot-tested in Ghana, launching Project Kané.^[7]

How it works

To make the user experience authentic and pleasant in assisted reading,^[8] a story can be selected by the child from a menu listing interesting stories from the Weekly Reader or another sources which also includes user authored stories. Further, a tutor listens to the children read aloud using Carnegie Mellon’s Sphinx – II Speech Recognizer^[9][10] to process and interpret the student's oral reading. When a student makes mistakes, gets stuck or clicks for help then the tutor intervenes in the process.^[11] Then, the tutor responds with assistance modeled in part after expert reading teachers adapted to the limitations and capability of technology.

The Reading Tutor dynamically updates the estimate data of student’s reading level, and picks stories a bit harder according to their level; this approach allows the tutor to focus on the zone of proximal development. It also scaffolds key processes in reading. It explains unfamiliar word sand concepts by presenting short factoids (comparison to other words). It also provides spoken and graphical assistance when it notices student click for help, get stuck, skip a word, misread word or make a mistake. Visual speech of the tutor makes use of talking mouth video clips of phonemes. It assists word identification by previewing new words and reading hard words aloud. It also motivates students by praising their good performance.^[12]

Prototype Testing

Project has demonstrated usability, user recognition and acceptance, assistive effectiveness and even pre-to post-test gains. Successive versions of reading tutor with minimal use of 20 minutes per day has produced higher comprehension gains than current methods in controlled studies over several months. To ensure there was no third variable involved, different treatments were compared within the same classrooms and randomized treatment assignment, stratifying by pretest scores within class. Valid and reliable measures (Woodcock.1998)^[13] were used to measure gains between pre and post test.^[14]

Various Test Studies occurred over last decade:

1. Pilot Study(1996–97) ^[15]
2. Within-classroom comparison(1998) ^[16]
3. Comparison to human tutors(1999-2000) ^[17]
4. Equal-time comparison to Sustained Silent Reading (2000-2001) ^[18]

Since, 2005 many researchers have conducted and published controlled studies of the reading tutor.^[19]

Awards

Project Listen has received global recognition and many awards:

1. 2012: Best Student Paper Award at 5th International Conference on Educational Data Mining^[20]
2. 2012: Best Paper Award at 25th Florida Artificial Intelligence Research Society Conference (FLAIRS-25).^[21]
3. 2011: Best Poster Nominee at 4th International Conference on Educational Data Mining.^[22]
4. 2011: Best Paper Nominee at 15th International Conference on Artificial Intelligence in Education.^[23]
5. 2009: Honorable Mention Student Paper Award at 3rd IEEE/ACM International Conference on Information and Communication Technologies and Development.^[24]
6. 2008: Best Paper Award at the 9th International Conference on Intelligent Tutoring Systems.^[25]
7. 2008: Best Paper Nominee at the 9th International Conference on Intelligent Tutoring Systems^[26]
8. 2002: Included in PBS series Reading Rockets^[27]
9. 2002: Distinguished Finalist for the International Reading Association's Outstanding Dissertation of the Year Award ^[28]
10. 2000: Included in National Science Foundation's "Nifty 50"^[29]
11. 1998: Represented the Computing Research Association (CRA) at the Coalition for National Science Funding Exhibit (CNSF) for Congress
12. 1994: Outstanding Paper Award at the Twelfth National Conference on Artificial Intelligence^[30]

See also

• Speech recognition
• Artificial Intelligence
• Educational data mining
• Learning Sciences

References

External links

• http://www.cs.cmu.edu/~listen/