Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Platform for Generating Educational Activities to Teach English as a Second Language

Aiala Rosá, Santiago Góngora, Juan Pablo Filevich, Ignacio Sastre, Laura Musto, Brian Carpenter, Luis Chiruzzo

TL;DR

The paper presents a web-based platform for English as a second language education that combines semi-automatic vocabulary resources, an image bank, and a suite of activity generators (both out-of-the-box and text-driven) under a teacher-reviewed workflow. It emphasizes a unified data format and resource sharing across activities, enabling learners to engage through a variety of games and exercises. The work reflects five years of development in collaboration with Uruguay's National Public Education Administration, addressing deployment constraints on commodity servers and advocating teacher supervision when using neural models to mitigate content quality and safety concerns. It also reports initial forays into image- and text-generation capabilities, outlining practical challenges and a roadmap toward more powerful hosting and deeper integration of experimental tools to broaden ESL education.

Abstract

We present a platform for the generation of educational activities oriented to teaching English as a foreign language. The different activities -- games and language practice exercises -- are strongly based on Natural Language Processing techniques. The platform offers the possibility of playing out-of-the-box games, generated from resources created semi-automatically and then manually curated. It can also generate games or exercises of greater complexity from texts entered by teachers, providing a stage of review and edition of the generated content before use. As a way of expanding the variety of activities in the platform, we are currently experimenting with image and text generation. In order to integrate them and improve the performance of other neural tools already integrated, we are working on migrating the platform to a more powerful server. In this paper we describe the development of our platform and its deployment for end users, discussing the challenges faced and how we overcame them, and also detail our future work plans.

A Platform for Generating Educational Activities to Teach English as a Second Language

TL;DR

The paper presents a web-based platform for English as a second language education that combines semi-automatic vocabulary resources, an image bank, and a suite of activity generators (both out-of-the-box and text-driven) under a teacher-reviewed workflow. It emphasizes a unified data format and resource sharing across activities, enabling learners to engage through a variety of games and exercises. The work reflects five years of development in collaboration with Uruguay's National Public Education Administration, addressing deployment constraints on commodity servers and advocating teacher supervision when using neural models to mitigate content quality and safety concerns. It also reports initial forays into image- and text-generation capabilities, outlining practical challenges and a roadmap toward more powerful hosting and deeper integration of experimental tools to broaden ESL education.

Abstract

We present a platform for the generation of educational activities oriented to teaching English as a foreign language. The different activities -- games and language practice exercises -- are strongly based on Natural Language Processing techniques. The platform offers the possibility of playing out-of-the-box games, generated from resources created semi-automatically and then manually curated. It can also generate games or exercises of greater complexity from texts entered by teachers, providing a stage of review and edition of the generated content before use. As a way of expanding the variety of activities in the platform, we are currently experimenting with image and text generation. In order to integrate them and improve the performance of other neural tools already integrated, we are working on migrating the platform to a more powerful server. In this paper we describe the development of our platform and its deployment for end users, discussing the challenges faced and how we overcame them, and also detail our future work plans.
Paper Structure (12 sections, 5 figures)

This paper contains 12 sections, 5 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. The Platform
  4. Resources Shared by Different Activities
  5. Words and Definitions
  6. Using both Text and Images
  7. Generating Activities from Text
  8. Image and Text Generation
  9. Challenges and Limitations
  10. Challenges faced in the development and deployment of the platform
  11. Using neural models in teaching environments
  12. Conclusions

Figures (5)

  • Figure 1: Screenshot of the landing screen of the platform.
  • Figure 2: A game consisting of choosing the image related to the word "tennis".
  • Figure 3: Story game based on the story "The fox and the grapes" with sentences shuffled, the student must put them in the correct order.
  • Figure 4: (A) Resulting image for "At the top of the wall, he saw the biggest, juiciest grapes he'd ever seen". (B) Resulting image with the co-reference resolution, changing the pronoun he for a fox.
  • Figure 5: Prompt example for the generation of a sentence using GPT-2.