Learning Student and Content Embeddings for Personalized Lesson Sequence Recommendation

Reddy, Siddharth; Labutov, Igor; Joachims, Thorsten

doi:10.1145/2876034.2893375

Cited by 22 publications

(19 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In order to learn non-trivial navigational patterns from past course events, we also wanted a course with a high amount of variation in navigational pathways exhibited by its learners. To measure this variation, we chose to treat student paths through a particular course as a Markov chain and then computed the entropy of the transition probability matrix for each course [26]. There were 13 courses evaluated offered by our deployment University partner, DelftX.…”

Section: Choice Of Coursementioning

confidence: 99%

Enabling Real-Time Adaptivity in MOOCs with a Personalized Next-Step Recommendation Framework

Pardos

Tang

Davis

et al. 2017

Proceedings of the Fourth (2017) ACM Conference on Learning @ Scale

View full text Add to dashboard Cite

In this paper, we demonstrate a first-of-its-kind adaptive intervention in a MOOC utilizing real-time clickstream data and a novel machine learned model of behavior. We detail how we augmented the edX platform with the capabilities necessary to support this type of intervention which required both tracking learners' behaviors in real-time and dynamically adapting content based on each learner's individual clickstream history. Our chosen pilot intervention was in the category of adaptive pathways and courseware and took the form of a navigational suggestion appearing at the bottom of every non-forum content page in the course. We designed our pilot intervention to help students more efficiently navigate their way through a MOOC by predicting the next page they were likely to spend significant time on and allowing them to jump directly to that page. While interventions which attempt to optimize for learner achievement are candidates for this adaptive framework, behavior prediction has the benefit of not requiring causal assumptions to be made in its suggestions. We present a novel extension of a behavioral model that takes into account students' time spent on pages and forecasts the same. Several approaches to representing time using Recurrent Neural Networks are evaluated and compared to baselines without time, including a basic n-gram model. Finally, we discuss design considerations and handling of edge cases for real-time deployment, including considerations for training a machine learned model on a previous offering of a course for use in a subsequent offering where courseware may have changed. This work opens the door to broad experimentation with adaptivity and serves as a first example of delivering a data-driven personalized learning experience in a MOOC.

show abstract

Section: Choice Of Coursementioning

confidence: 99%

Enabling Real-Time Adaptivity in MOOCs with a Personalized Next-Step Recommendation Framework

Pardos

Tang

Davis

et al. 2017

Proceedings of the Fourth (2017) ACM Conference on Learning @ Scale

View full text Add to dashboard Cite

show abstract

“…Embeddings are dense, lower-dimensional representations that are derived from sequentially ordered data and encode transition probabilities based on the observations in the original data. They were applied, for example, in the domains of next-track music recommendation [Chen et al 2012;Zheleva et al 2010], recommendation of learning courses [Reddy et al 2016], or next point-of-interest (POI) recommendation [Feng et al 2015]. However, a general challenge when using item embeddings is that they can be computationally demanding and sometimes require substantial amounts of training data to be effective.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of session-based recommendation algorithms

Ludewig

Jannach

2018

User Model User-Adap Inter

260

278

View full text Add to dashboard Cite

Recommender systems help users find relevant items of interest, for example on e-commerce or media streaming sites. Most academic research is concerned with approaches that personalize the recommendations according to long-term user profiles. In many real-world applications, however, such long-term profiles often do not exist and recommendations therefore have to be made solely based on the observed behavior of a user during an ongoing session. Given the high practical relevance of the problem, an increased interest in this problem can be observed in recent years, leading to a number of proposals for session-based recommendation algorithms that typically aim to predict the user's immediate next actions.In this work, we present the results of an in-depth performance comparison of a number of such algorithms, using a variety of datasets and evaluation measures. Our comparison includes the most recent approaches based on recurrent neural networks like gru4rec, factorized Markov model approaches such as fism or fossil, as well as simpler methods based, e.g., on nearest neighbor schemes. Our experiments reveal that algorithms of this latter class, despite their sometimes almost trivial nature, often perform equally well or significantly better than today's more complex approaches based on deep neural networks. Our results therefore suggest that there is substantial room for improvement regarding the development of more sophisticated session-based recommendation algorithms. 1

show abstract

“…On the other hand, implicitly gathered information about the user, which does not require the user's cooperation, will give the system more accurate information about the user's interest, how the user uses the system or what contents are recommended, etc. (Reddy, 2016). However, for implicit data collection, the user must use the system for a certain period of time.…”

Section: Common Problems In Educational Recommender Systemsmentioning

confidence: 99%

Educational Recommender Systems: An Overview and Guidelines for Further Research and Development / Obrazovni sustavi preporučivanja: pregled stanja sa smjernicama za daljnja istraživanja i razvoj

Đurović

Dlab

Hoić-Božić

2018

Croatian Journal of Education

View full text Add to dashboard Cite

Educational Recommender Systems (ERS) are increasingly used as tools to help students and teachers during the implementation of the learning process. The main difference between ERS and their commercial counterparts is in the pedagogical principles appropriate for the learning and teaching process. The differences in the educational methods used in a variety of educational situations, and their dependence on the field of study, set initial guidelines for ERS design. This paper reviews the evolution of ERS up to the currently achieved level of development and presents the basic techniques used in ERS design and the common problems they encounter in their work. Examples of classification of different ERS, according to their specific characteristics and basic approaches in their work, are presented. Based on this analysis, along with the training and upgrading of the existing algorithms, five specific areas in which future research and development can be expected are defined: construction of universal ERS, ERS intended primarily for teachers, ERS that links student achievements across different courses, ERS which take into account physical distance between students and use of ERS to motivate students to work continuously.

show abstract

Learning Student and Content Embeddings for Personalized Lesson Sequence Recommendation

Cited by 22 publications

References 19 publications

Enabling Real-Time Adaptivity in MOOCs with a Personalized Next-Step Recommendation Framework

Enabling Real-Time Adaptivity in MOOCs with a Personalized Next-Step Recommendation Framework

Evaluation of session-based recommendation algorithms

Educational Recommender Systems: An Overview and Guidelines for Further Research and Development / Obrazovni sustavi preporučivanja: pregled stanja sa smjernicama za daljnja istraživanja i razvoj

Contact Info

Product

Resources

About