HP Labs India

Future School: Learning without Constraints

The rapid growth of the Internet has created a medium where education can be delivered over the web in an engaging, broad reaching, cost-effective and 24x7 manner. The goal of this project is to create "Future School", a cloud based learning platform that can allow personalized education delivery over the Internet in the most intuitive and engaging manner. We envision a system that will allow anyone with a computing device and internet connection in any part of the world to access quality education at an affordable cost.

Motivation

Current methods of education delivery are centuries old, not easily scalable, and require significant physical infrastructure (e.g., roads, transportation, classrooms, teaching staff on payrolls, paper textbooks). Brick-and-mortar schools restrict peer-interactions, while rigid curricula constrain student learning. The goal of Future School is to bring together research and advanced development with a single purpose: to create an engaging and personalized cloud-based learning experience.

Research Threads

The core challenges in Future School require us to take a comprehensive approach, drawing from diverse fields like machine learning and analytics, information retrieval, natural language processing and information theory, with the aim of building scalable cloud-based solutions. Some of the key research challenges are:

  • Providing an engaging online learning experience: Worldwide, textbooks are predicted to go digital from 3% today to 25% in 2015. However, student learning still remains static, and does not utilize the digital content. There is also a tremendous amount of content on the Internet that is relevant and useful to a student. The challenge is to bring these together in a suitable form so as to enhance student learning. Research in text/video processing, automated content creation, translation, summarization, ranking, recommender systems etc. are key enablers.
     
  • Personalized and Collaborative learning: One-size-fits-all teaching methods are neither effective nor acceptable for today's diverse student population. We model a student's learning activity to track and predict student engagement, performance and progress. By using automated assessment along with feedback-based learning and optimization techniques, we are aiming to fully personalize the learning experience. Research in interactive and group learning is helping us uncover the potential of peer collaborations.
     
  • Large-scale platforms for education delivery: Today, there is no single large platform that provides content consumption and assessment under one roof. The challenge is to create a large-scale platform that can gather data as well as analytics that can be used to truly personalize learning and assessment, and provide a collaborative learning experience. Component technologies include Mapreduce frameworks (e.g. Hadoop), Graph databases (e.g. Neo4J), big graph processing (e.g. Giraph), large scale machine learning (e.g. Mahout) etc.
Awards
  • VideoBook from HP Labs India has won the 2012 Wall Street Journal Technology Innovation Award in the Education Technology category.
  • Future School has won the Technology Review India's 2010 Grand Challenges for Technologists (2010 TRGC) Award.
Relevant Publications
Other links
Research Team

For further information or potential career opportunities in Future School please contact Krishnan Ramanathan.

This page was last updated on February 13, 2013