Data Innovators Knewton's logo

Published on January 31st, 2014 | by Travis Korte

0

5 Q’s for Knewton Data Science Director George Davis

The Center for Data Innovation spoke with George Davis, Director of Data Science at New York-based personalized education startup Knewton, about how machine learning can help design a more tailored curriculum for students in K-12, higher education, corporate training and other environments.

This interview has been lightly edited.

Travis Korte: For those who may be unfamiliar, can you give a brief introduction to Knewton, what you make, and who uses it?

George Davis: Knewton personalizes digital courses so every student is engaged and no student slips through the cracks.

Knewton provides an infrastructure platform that allows others to build powerful proficiency-based adaptive learning applications. Each third-party app Knewton powers brings its own core competencies in the subtle arts of content creation, pedagogy, and user experience, while outsourcing the heavy machinery of its personalization infrastructure to Knewton. The world’s top educational publishers (and soon anyone who creates lessons) can use Knewton to improve learning outcomes in K–12, higher education, English-language teaching, corporate training, and other markets.

As students work through online lessons, Knewton analyzes vast amounts of anonymized data to figure out what a student knows and how they learn best. Then Knewton recommends what to study next, helping students at any level succeed. Teachers use Knewton-powered real-time predictive analytics to detect gaps in knowledge and differentiate instruction. Students who have different needs, interests, strengths, and weaknesses can work toward goals in a sequence and pace that continuously adjusts to fit their needs.

TK: What is “adaptive learning?” What does it do better than traditional teaching methods?

GD: Truly adaptive learning technology can provide continuously optimized recommendations to students at any given moment. While it’s relatively straightforward to make simple differentiated learning apps, it’s extremely difficult to make proficiency-based adaptive learning. A true model of proficiency can estimate what students know, how prepared they are for further instruction or assessment, and how their abilities evolve over time.

Knewton is often confused with the many learning apps currently on the market. However, Knewton isn’t an app or just one adaptive course or lesson—it’s an infrastructure platform that uses concept-level proficiency data to personalize content and learning materials that others create.

As a student progresses through a digital course (for example, whenever a student watches a video, reads a lesson, answers a practice problem, or takes a quiz), anonymized data on what they know and the effectiveness of the content are analyzed in aggregate with millions of other anonymized students.

Knewton’s machine learning algorithms observe patterns and propose underlying factors that help explain the data received. Material that has been consistently helpful to students with a particular challenge—say, difficulty understanding the order of mathematical operations—can be routed to a student who is having trouble with assessments of a post-requisite skill. By measuring student sensitivities to content properties such as presentation modality or teaching strategy, we can prioritize content that is both relevant to their needs and able to keep them engaged longer.

Knewton also gives teachers analytics and tools to help educators tailor their lessons or differentiate instruction depending on what the class or a particular student needs. Knewton helps teachers monitor performance and reduce administrative work, giving them more time to do what they do best—teach and inspire students.

TK: What new insights or initiatives would be possible with widespread adoption of adaptive learning technologies in, say, K-12 education?

GD: It’s interesting to think about the cross-disciplinary implications. As students develop a robust learning history, Knewton can also help connect a student’s various areas of coursework. Since each subject in a traditional high school or middle school requires a different teacher with the correct area of expertise, various subjects are often presented to students as being far more distinct and separate from each other than they actually are.

Knewton uses sophisticated “knowledge graphs”—graphs of academic concepts, linked by prerequisite relationships that help define a student’s path through courses—that can link cross-disciplinary material. For example, Knewton could help discover that a student who is weak with math word problems is actually struggling because of reading comprehension difficulties; Knewton may then suggest that student work on material that improves grasp of syntax and vocabulary.

TK: Can you speak about some interesting applications that have already developed using your platform?

GD: Top publishers around the globe are integrating the Knewton API into their digital courses and apps to provide continuously personalized learning. Knewton helps publishers deliver differentiated instruction for students from preschool to higher education, adult learning, and corporate training. Knewton personalizes learning for subjects including math, sciences, reading, writing, business, language learning, sociology, and more.

Many of the world’s top educational publishers, including Pearson, Cambridge University Press, Cengage Learning, Houghton Mifflin Harcourt, Macmillan Education, and more, have already partnered with us to create Knewton-powered lessons and courses.

While it will take time to demonstrate long-term outcomes, early evidence points to Knewton’s efficacy. Since math students at Arizona State University began using Knewton-powered courses, pass rates have risen from 64% to 75% and withdrawal rates have been cut in half. Almost half of students taking Knewton-powered math courses finished class four weeks early.

TK: The benefit of adaptive learning for students seems apparent, but can you speak a little about the benefits education providers can reap from such technologies?

GD: In addition to providing personalized recommendations for students and empowering educators with useful information, we can help education providers create more effective lessons and learning apps. Knewton can analyze which lessons resonate best, for whom, and why—enabling publishers and content creators to evaluate content efficacy at the concept level and create more effective learning materials.

Tags: , , , , , , , , , , ,


About the Author

Travis Korte is a research analyst at the Center for Data Innovation specializing in data science applications and open data. He has a background in journalism, computer science and statistics. Prior to joining the Center for Data Innovation, he launched the Science vertical of The Huffington Post and served as its Associate Editor, covering a wide range of science and technology topics. He has worked on data science projects with HuffPost and other organizations. Before this, he graduated with highest honors from the University of California, Berkeley, having studied critical theory and completed coursework in computer science and economics. His research interests are in computational social science and using data to engage with complex social systems. You can follow him on Twitter @traviskorte.



Back to Top ↑
  • Sign up for our weekly newsletter