[TMP-039] Educational Recommenders with Narratives

Contact person: John Shawe-Taylor, UCL (j.shawe-taylor@ucl.ac.uk) 

Internal Partners:

1. University College London í(UCL), John Shawe-Taylor
2. Institut “Jožef Stefan” (JSI), John Shawe-Taylor
3. INESC TEC, Alipio Jorge  

Through this work, we explore novel and advanced learner representation models aimed at exploiting learning trajectories to build a transparent, personalised and efficient automatic learning tutor through resource recommendations. We elaborate on the different types of publicly available data sources that can be used to build an accurate trajectory graph of how knowledge should be taught to learners to fulfil their learning goals effectively. Our aim is to capture and utilise the inferred learner state and the understanding the model has about sensible learning trajectories to generate personalised narratives that will allow the system to rationalise the educational recommendations provided to individual learners. Since an educational path consists heavily of building/following a narrative, a properly constructed narrative structure and representation is paramount to the problem of building successful and transparent educational recommenders.

Results Summary

Adding humanly-intuitive model assumptions to the TrueLearn Bayesian learner model such as 1) interest, 2) knowledge of learner 3) semantic relatedness between content topics has been achieved successfully leading to improved predictive performance. A dataset of personalised learning pathways of over 20000 learners has been composed. Analysis on Optimal Transport for generating interpretable narratives using Earth Mover’s Distance (EMD) of Wikipedia concepts also showed promise in scenarios where there is a limited number of topic annotations per document. A novel method for cross-lingual information retrieval using EMD has been invented pursuing this idea. Incorporating semantic networks (WordNet, WikiData) in building higher-level reasoning for recommendation also shows promise albeit with limited results at this point. Successful expansion of WordNet network using WikiData network is achieved. The resultant semantic network indicates that the quality of reasoning over Wiki Annotated video lectures can be improved in this way.

Tangible Outcomes

1. X5Learn: A Personalised Learning Companion at the Intersection of AI and HCI – Maria Perez-Ortiz. https://dl.acm.org/doi/10.1145/3397482.3450721  
2. “Why is a document relevant? Understanding the relevance scores in cross-lingual document retrieval.” Novak, Erik, Luka Bizjak, Dunja Mladenić, and Marko Grobelnik. Knowledge-Based Systems 244 (2022): 108545. https://dl.acm.org/doi/10.1016/j.knosys.2022.108545
3. Towards an Integrative Educational Recommender for Lifelong Learners (Student Abstract). Sahan Bulathwela,  María Pérez-Ortiz, Emine Yilmaz, and John Shawe-Taylor. (2020, April). In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 34, No. 10, pp. 13759-13760). https://ojs.aaai.org/index.php/AAAI/article/view/7151/7005
4. “TrueLearn: A Python Library for Personalised Informational Recommendations with (Implicit) Feedback.” Yuxiang Qiu,  Karim Djemili, Denis Elezi, Aaneel Shalman, María Pérez-Ortiz, and Sahan Bulathwela.  arXiv preprint arXiv:2309.11527 (2023). Published through ORSUM workshop, RecSys’23 https://arxiv.org/pdf/2309.11527
5. “Peek: A large dataset of learner engagement with educational videos.” Bulathwela, Sahan, Maria Perez-Ortiz, Erik Novak, Emine Yilmaz, and John Shawe-Taylor.  arXiv preprint arXiv:2109.03154 (2021). Submitted to ORSUM workshop, RecSys’21 https://arxiv.org/abs/2109.03154
6. Dataset: PEEK Dataset – Sahan Bulathwela https://github.com/sahanbull/PEEK-Dataset
7. Program/code: TrueLearn Model – Sahan Bulathwela https://github.com/sahanbull/TrueLearn
8. Program/code: Semantic Networks for Narratives – Daniel Loureiro https://github.com/danlou/mp_narrative