# T-TExTS (Teaching Text Expansion for Teacher Scaffolding): Enhancing Text Selection in High School Literature through Knowledge Graph-Based Recommendation

arXiv: https://arxiv.org/abs/2506.12075 · HTML: https://mlsanity.com/paper/2506.12075/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 60 references that are in the map.
- First posted on arXiv 6 Jun 2025.

## Builds on (most cited first, 21 of 21)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Efficient Estimation of Word Representations in Vector Space](https://mlsanity.com/paper/1301.3781/) (2013)
- [node2vec: Scalable Feature Learning for Networks](https://mlsanity.com/paper/1607.00653/) (2016)
- [DeepWalk: Online Learning of Social Representations](https://mlsanity.com/paper/1403.6652/) (2014)
- [From Local to Global: A Graph RAG Approach to Query-Focused Summarization](https://mlsanity.com/paper/2404.16130/) (2024)
- [Embedding Entities and Relations for Learning and Inference in Knowledge Bases](https://mlsanity.com/paper/1412.6575/) (2014)
- [Benchmarking Graph Neural Networks](https://mlsanity.com/paper/2003.00982/) (2020)
- [Knowledge Graphs](https://mlsanity.com/paper/2003.02320/) (2020)
- [A Survey on Knowledge Graph-Based Recommender Systems](https://mlsanity.com/paper/2003.00911/) (2020)
- [Knowledge Graph Embedding for Link Prediction: A Comparative Analysis](https://mlsanity.com/paper/2002.00819/) (2020)
- [Graph Representation Learning: A Survey](https://mlsanity.com/paper/1909.00958/) (2019)
- [Knowledge Graph Contrastive Learning for Recommendation](https://mlsanity.com/paper/2205.00976/) (2022)
- [A Survey on Graph Representation Learning Methods](https://mlsanity.com/paper/2204.01855/) (2022)
- [OWL2Vec*: Embedding of OWL Ontologies](https://mlsanity.com/paper/2009.14654/) (2020)
- [Neuro-symbolic representation learning on biological knowledge graphs](https://mlsanity.com/paper/1612.04256/) (2016)
- [OPA2Vec: combining formal and informal content of biomedical ontologies to improve similarity-based prediction](https://mlsanity.com/paper/1804.10922/) (2018)
- [Onto2Vec: joint vector-based representation of biological entities and their ontology-based annotations](https://mlsanity.com/paper/1802.00864/) (2018)
- [Personalized Audiobook Recommendations at Spotify Through Graph Neural Networks](https://mlsanity.com/paper/2403.05185/) (2024)
- [Negative Sampling in Knowledge Graph Representation Learning: A Review](https://mlsanity.com/paper/2402.19195/) (2024)
- [Fast and scalable learning of neuro-symbolic representations of biomedical knowledge](https://mlsanity.com/paper/1804.11105/) (2018)

## Built on by (most cited first, 1 of 1)

- [LIT-GRAPH: Evaluating Deep vs. Shallow Graph Embeddings for High-Quality Text Recommendation in Domain-Specific Knowledge Graphs](https://mlsanity.com/paper/2602.07307/) (2026)

## Abstract (arXiv)

High school English Literature teachers often encounter barriers to assembling diverse, thematically aligned text sets due to limited planning time and pedagogical resources. To address this need, we present T-TExTS (Teaching Text Expansion for Teacher Scaffolding), a knowledge graph (KG)-based recommendation system that suggests literature texts based on pedagogical merit rather than surface-level metadata. We construct a domain-specific ontology using the Knowledge Acquisition and Representation Methodology (KNARM), instantiate it as a knowledge graph with separate Terminological Box (TBox) and Assertional Box (ABox) components, and evaluate four graph embedding strategies (DeepWalk, biased random walk, hybrid embedding, and Node2Vec) across three dataset configurations (98, 196, and 351 texts) and two relation-weighting schemes. The experimental results reveal that traversal-level expert weighting alone does not outperform algorithmic structural tuning: Node2Vec achieves the highest Area Under the Curve (AUC) at every dataset size (0.9642--0.9750) and the strongest ranking metrics (Hits@K, MRR, nDCG) at larger scales. Combining structural and pedagogical signals through embedding concatenation, however, preserves both interpretability and competitive ranking quality, with the hybrid model maintaining a high AUC across all scales (0.9122--0.9350) and remaining within a few percentage points of Node2Vec on every ranking metric. These findings highlight the value of ontology-driven knowledge graph embeddings for educational recommendation systems and demonstrate that T-TExTS can meaningfully ease the burden of English Literature text selection for secondary educators, supporting more informed and inclusive curricular decisions. The source code for T-TExTS is available at https://github.com/koncordantlab/TTExTS.