Neural Educational Recommendation Engine (NERE)

arXiv.org Machine Learning

Quizlet is the most popular online learning tool in the United States, and is used by over 2/3 of high school students, and 1/2 of college students. With more than 95% of Quizlet users reporting improved grades as a result, the platform has become the de-facto tool used in millions of classrooms. In this paper, we explore the task of recommending suitable content for a student to study, given their prior interests, as well as what their peers are studying. We propose a novel approach, i.e. Neural Educational Recommendation Engine (NERE), to recommend educational content by leveraging student behaviors rather than ratings. We have found that this approach better captures social factors that are more aligned with learning. NERE is based on a recurrent neural network that includes collaborative and content-based approaches for recommendation, and takes into account any particular student's speed, mastery, and experience to recommend the appropriate task. We train NERE by jointly learning the user embeddings and content embeddings, and attempt to predict the content embedding for the final timestamp. We also develop a confidence estimator for our neural network, which is a crucial requirement for productionizing this model. We apply NERE to Quizlet's proprietary dataset, and present our results. We achieved an R^2 score of 0.81 in the content embedding space, and a recall score of 54% on our 100 nearest neighbors. This vastly exceeds the recall@100 score of 12% that a standard matrix-factorization approach provides. We conclude with a discussion on how NERE will be deployed, and position our work as one of the first educational recommender systems for the K-12 space.


Artificial intelligence has a racial bias problem. Google is funding summer camps to try to change that

USATODAY

On a sunny Monday afternoon in Oakland, AI4All alum Ananya Karthik gathered a few dozen girls to show them how to use the Deep Dream Generator program to fuse images together and create a unique piece of art. OAKLAND -- Through connections made at summer camp, high school students Aarvu Gupta and Lili Sun used artificial intelligence to create a drone program that aims to detect wildfires before they spread too far. Rebekah Agwunobi, a rising high school senior, learned enough to nab an internship at the Massachusetts Institute of Technology's Media Lab, working on using artificial intelligence to evaluate the court system, including collecting data on how judges set bail. Both projects stemmed from the Oakland, Calif.-based nonprofit AI4All, which will expand its outreach to young under-represented minorities and women with a $1 million grant from Google.org, the technology giant's philanthropic arm announced Friday. Artificial intelligence is becoming increasingly commonplace in daily life, found in everything from Facebook's face detection feature for photos to Apple's iPhone X facial recognition.


Mandan Schools Work to Get All Students IPads

U.S. News

The district has been moving to integrate iPads in classrooms as part of Ignite 2025, a plan for equity to ensure all students have access to technology. The district purchased about 900 devices for Mandan Middle School students last year and 1,000 devices for high school students the year before.


Have You Ever…

Slate

Once upon a teenager, on a faraway internet--years before Facebook chain letters and BuzzFeed's revamp of the age-old personality quiz--a silly questionnaire swept the nation. A kind of standardized version of "Never Have I Ever," the quiz made its way from college dorms to high-school LiveJournal entries to middle-school bus trips, measuring sexual and criminal deviance on a 100-point scale. It went viral in the peculiar ways of the late 1990s and early 2000s, after the advent of the internet but before the explosion of social media. It was called the Purity Test.


For principals, student sexting a speeding 'freight train,' full of peril

PBS NewsHour

For Jane Griffin, the principal at Louisiana's Winnfield High, the moment came when one of her students found a staff member's smartphone laying on a desk, picked it up, and took a picture of his own genitals. For Shafta Collazo, an assistant principal at Delaware's Woodbridge Middle School, it came when a student got mad at his girlfriend and decided to "Airdrop" compromising digital photos of her to dozens of other children using a file-transfer service for Mac devices. And for assistant principal Deirdra Chandler, the harsh realization that responding to youth "sexting" is now an inescapable part of the job, even for leaders of kindergarten through fifth-grade schools, came after one of her young students at South Carolina's Erwin Elementary sent out sexual imagery of another student to his friends. "It's scary," said Chandler, one of nearly 100 concerned school leaders who packed into a room at the annual conference of the National Association of Secondary School Principals and National Association of Elementary School Principals in Philadelphia on July 10, to discuss the dangers of sexting. This fraught new reality for U.S. schools is regularly in the headlines -- recently, the Providence Journal ran a story about an "online school Dropbox" full of graphic sexual images and videos of dozens of mostly underage female students at a local high school.