REAP.PT - Computer Aided Language Learning (CALL) Reading Practice (REAP.PT)
Start Date: April 1st, 2009 End Date: October 30th, 2012
PIs: Nuno Mamede (IST/UTL and INESC ID) and Maxine Eskenazi (CMU)
Co-PIs: Jorge Baptista (UAlg) and Maria do Céu Ribeiro (FUL/UL)
Dual Degree Ph.D. Students: Luís Carlos dos Santos Marujo (Language Technologies), Rui Pedro dos Santos Correia (Language Technologies)
Teams: INESC-ID Lisboa, Universidade do Algarve, Universidade de Lisboa, Carnegie Mellon University
Companies: Porto Editora
In order to enable students to learn to read another language, a good tutoring system should give them much opportunity for practice and make the experience as engaging and personalized as possible. The REAP.PT system is being designed to complement teacher time by giving the student documents to read and questions about new words they have seen in the documents. It will personalize the work by choosing texts in Portuguese that are at the reading level of the individual student, presenting words that that student needs to learn and having documents on subjects that the student is interested in. Questions will be automatically generated about the meaning of the words that the student saw in a document and reports will be given to the student and to their teacher.
One challenge is how to find enough documents to satisfy all of the different students’ needs. Portuguese documents are not as numerous as English ones on the Web and so we need to find other sources of documents, such as from publishers. Another challenge is how to automatically determine the level of a document. Here we need to bootstrap from a set of documents with predetermined levels and a word and grammar‐based analysis. Once we have a reading curriculum set up, another challenge concerns maintaining student interest.
While reading documents, and thus learning words from context, provides the basic format that is best for learning new words, the number of words that can be learned in one session is limited. We also believe that students expect more from a tutoring system in the twenty first century. They expect it to be entertaining. Games have been shown to increase interest, motivation and learning. We will therefore create a suite of games for REAP that enable learning of more vocabulary than can be learned in classroom time. The issue here is to find the correct educational game form or forms (teaches the material we want to teach, motivates the student to spend a lot of time on it, promotes learning), to create the databases of curricular materials that the student players need to learn, to design and implement the interface that presents the game (web‐enabled), and to create an intelligent tutor that constantly tallies correct and incorrect answers in order to keep the next question challenging and engaging.