Create a self-adapting learning system that can scale to millions of users
Use Elasticsearch to search, index, and rank content recommendations from the virtually unlimited data on the Internet
Case Study Highlights
Create an amazing user experience
- Index and recommend virtually unlimited amounts of content
- Recommend content based on user preference and content popularity
- Deliver context-specific results in real time
- Develop and scale easily
Develop and scale easily
- Easily search structured and unstructured data
- Make new content searchable immediately
- Provision new nodes automatically
"We're creating a consumer search and ranking service, and our goal was how to make that scale. A lot of solutions were too heavy or slow. Once we found Elasticsearch, the team made a quick decision to use it."
Learning is moving online, whether it is through stand-alone curriculum or a hybrid online/classroom environment. McGraw-Hill has been responding to the growing demand for technology in the classroom by moving from a textbook business model to a digital-based model. Their focus in recent years has been on developing adaptive learning systems that enable classroom teaching to come closer to a one-to-one student-teacher interaction. “Gone are the days when one teacher delivers the same lesson plan to 30 students," explains Chris Tse, Head of R&D at McGraw-Hill Education Labs. “Today, instructors are using technology to customize content, learning, and even tests to a student's preferences and learning style."
Chris had used MarkLogic extensively in his previous role at BusinessWeek, so he began evaluating MarkLogic for the EdSense platform. “It was clear that it was going to be a struggle for MarkLogic to scale to a consumer-grade service," he explains. “You quickly run into how heavy it is as well as its latency issues." It was then that Chris discovered Elasticsearch. It didn't take long for Chris and his team to see that Elasticsearch could provide huge scalability with the high performance they needed.
Content discovery and recommendation for virtually unlimited content from the Internet
McGraw-Hill's role is changing from creating textbook content to delivering third party content in a usable format. Elasticsearch enables them to scale the amount of content they are able to index, search, and recommend, which in turn scales the value of their platform.
Monitor and monetize third party content
EdSense searches and recommends content from a myriad of third party providers, much of which is paid content. Elasticsearch allows EdSense to monitor, bill, and pay for usage of this material.
Scale distribution to millions of learners
As online learning becomes more a part of everyday life, the company sees the EdSense opportunity to deliver adaptive learning to millions of people. They can scale to this magnitude easily with Elasticsearch.
Deliver content based on user preference and content popularity
Content is delivered in real-time to students based on their preferences (print vs. video) or previous behavior (a low score on a test). Using Elasticsearch, Edsense can deliver personalized student learning by assessing each student's skill level and using data to determine how each can progress through lessons most effectively.