pith. sign in

arxiv: 2506.12075 · v2 · pith:HGLKLCJXnew · submitted 2025-06-06 · 💻 cs.IR · cs.AI

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

Nirmal Gelal , Chloe Snow , Ambyr Rios , Kathleen M. Jagodnik , Hande K\"u\c{c}\"uk McGinty This is my paper

Pith reviewed 2026-05-19 10:29 UTC · model grok-4.3

classification 💻 cs.IR cs.AI
keywords knowledge graphrecommendation systemgraph embeddingsNode2Veceducational technologyhigh school literaturepedagogical ontologytext selection
0
0 comments X

The pith

A knowledge graph with pedagogical relations lets Node2Vec embeddings recommend high school literature texts by teaching merit rather than metadata.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

High school English teachers face tight planning time when building diverse and thematically connected text sets. The paper introduces T-TExTS, which first builds a domain-specific ontology with the KNARM method, turns it into a knowledge graph that separates general concepts from specific text instances, and then tests four embedding strategies on datasets of 98 to 351 texts. Node2Vec produces the highest recommendation accuracy at every scale, while a hybrid embedding that joins structural and expert-weighted signals stays close in performance and keeps the reasons for each suggestion traceable. If these results hold, the approach gives teachers a practical tool for informed, inclusive text choices without exhaustive manual search.

Core claim

T-TExTS instantiates a KNARM-derived ontology as a knowledge graph with TBox and ABox layers, then evaluates DeepWalk, biased random walk, hybrid, and Node2Vec embeddings under two weighting schemes. Node2Vec records the highest AUC (0.9642 to 0.9750) across all three dataset sizes and the best ranking metrics at larger scales; the hybrid model that concatenates structural and pedagogical signals keeps AUC between 0.9122 and 0.9350 while remaining within a few points of Node2Vec on Hits@K, MRR, and nDCG, showing that structural tuning can outperform simple expert weighting alone.

What carries the argument

The Node2Vec embedding computed on the knowledge graph whose relations encode pedagogical merit of literature texts.

If this is right

  • Larger text collections improve ranking quality when structural tuning such as Node2Vec is used.
  • Concatenating structural and pedagogical embeddings keeps recommendation accuracy competitive while letting users trace suggestions to specific relations.
  • Traversal-level expert weighting by itself does not surpass algorithmic structural adjustments.
  • The system reduces reliance on surface metadata for assembling thematically aligned text sets.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same ontology-plus-embedding pattern could be reused for text or resource recommendation in other subjects such as history or science.
  • Direct comparison of the system's suggestions against teacher-chosen sets or measured student outcomes would test real-world impact.
  • Extending the graph with student learning goals or demographic factors might allow more personalized recommendations.

Load-bearing premise

The ontology's entities and relations correctly capture what makes a text pedagogically valuable for high school English classes.

What would settle it

An experiment in which independent high school English teachers rate the recommended texts for actual classroom alignment and compare those ratings to the system's ranking scores.

read the original abstract

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.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The paper presents T-TExTS, a knowledge graph-based recommendation system for high school English Literature text selection. It constructs a domain-specific ontology via the KNARM methodology with TBox and ABox components to encode pedagogical merit, instantiates it as a KG, and evaluates four embedding strategies (DeepWalk, biased random walk, hybrid concatenation, and Node2Vec) across three dataset sizes (98, 196, 351 texts) and two relation-weighting schemes. Experiments report that Node2Vec achieves the highest AUC (0.9642–0.9750) and strongest ranking metrics (Hits@K, MRR, nDCG) at larger scales, while the hybrid embedding maintains competitive performance (AUC 0.9122–0.9350) with added interpretability. The work concludes that such embeddings can ease teacher burden in curricular decisions, with source code released.

Significance. If the central premise holds, the results demonstrate the utility of ontology-driven KG embeddings for educational recommendation, showing that structural tuning (Node2Vec) outperforms simple expert weighting and that hybrid embeddings preserve both performance and interpretability. The open-source code and consistent reporting across dataset scales and embedding variants are strengths that support reproducibility and allow direct comparison in future work on KG-based educational systems.

major comments (1)
  1. Experimental results section (and abstract): The reported AUC values (0.9642–0.9750 for Node2Vec) and ranking metrics evaluate how faithfully the embeddings reconstruct held-out edges and relations from the KNARM-instantiated KG. These metrics therefore test internal consistency with the authors' chosen ontology and weighting scheme rather than independent evidence that the encoded relations and entities reflect pedagogical merit as judged by teachers or measured by student outcomes. This assumption is load-bearing for the claim that recommendations are 'based on pedagogical merit rather than surface-level metadata' and that the system 'can meaningfully ease the burden of English Literature text selection.'
minor comments (2)
  1. Abstract: Limited detail is given on the specific relations and entity types used to operationalize pedagogical merit within the KNARM ontology; expanding this would clarify how the TBox/ABox components differ from standard metadata-based approaches.
  2. Evaluation setup: The three dataset sizes and two relation-weighting schemes are clearly described, but the manuscript would benefit from explicit discussion of how the held-out edges were selected to avoid leakage from the ontology construction process.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the thoughtful and constructive feedback. We address the major comment point by point below, clarifying the role of the KNARM methodology in grounding the knowledge graph in pedagogical expertise while acknowledging the scope of our current evaluation.

read point-by-point responses

  1. Referee: Experimental results section (and abstract): The reported AUC values (0.9642–0.9750 for Node2Vec) and ranking metrics evaluate how faithfully the embeddings reconstruct held-out edges and relations from the KNARM-instantiated KG. These metrics therefore test internal consistency with the authors' chosen ontology and weighting scheme rather than independent evidence that the encoded relations and entities reflect pedagogical merit as judged by teachers or measured by student outcomes. This assumption is load-bearing for the claim that recommendations are 'based on pedagogical merit rather than surface-level metadata' and that the system 'can meaningfully ease the burden of English Literature text selection.'

    Authors: We appreciate the referee's precise characterization of the evaluation. The link-prediction metrics do measure reconstruction fidelity within the KNARM-derived knowledge graph. However, the graph itself is not an arbitrary structure: KNARM is an established iterative methodology that elicits and formalizes domain knowledge directly from subject-matter experts (here, high-school literature educators and curriculum designers). The resulting TBox defines pedagogically relevant concepts and relations (e.g., thematic scaffolding potential, grade-level alignment, diversity of perspectives), while the ABox populates concrete text instances according to those expert criteria. Consequently, the high AUC scores (0.9642–0.9750) demonstrate that Node2Vec and the hybrid embedding faithfully preserve expert-encoded pedagogical relationships rather than surface metadata. We agree that teacher judgment studies or student-outcome measures would provide stronger external corroboration and constitute a natural next step. In the revised manuscript we will (1) add an explicit paragraph in the Discussion section describing the expert-driven construction process and (2) include a dedicated Limitations subsection that notes the absence of direct user validation and highlights the open-source release as an invitation for such studies. revision: partial

Circularity Check

0 steps flagged

No circularity: empirical embedding evaluation on held-out KG edges

full rationale

The paper constructs a domain-specific ontology via KNARM, instantiates a KG (TBox + ABox), trains four embedding models, and reports standard link-prediction / ranking metrics (AUC 0.9642–0.9750, Hits@K, MRR, nDCG) on held-out edges from the same graph. These metrics measure reconstruction fidelity to the authors' supplied structure and weights; they do not reduce any equation or claim to a fitted parameter defined by the same authors, nor rely on a self-citation chain for a uniqueness theorem. No self-definitional relations, ansatz smuggling, or renaming of known results appear. The derivation chain (ontology → embeddings → held-out evaluation) is experimentally independent of its inputs and therefore self-contained.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 1 invented entities

The central claim rests on the assumption that the custom ontology faithfully represents pedagogical relationships; this is not derived from data but postulated via the KNARM methodology. No free parameters are explicitly fitted in the reported results beyond standard embedding hyperparameters. No new physical or mathematical entities are invented.

free parameters (2)
  • relation weighting scheme
    Two schemes tested; chosen to combine traversal-level expert input with structural signals.
  • dataset size
    Three configurations (98, 196, 351 texts) used to test scaling behavior.
axioms (1)
  • domain assumption The KNARM methodology produces an ontology whose relations capture pedagogically relevant connections between texts.
    Invoked when constructing the TBox and ABox components of the knowledge graph.
invented entities (1)
  • T-TExTS recommendation system no independent evidence
    purpose: To suggest literature texts based on graph embeddings of pedagogical merit.
    New named artifact introduced to solve the text-selection problem.

pith-pipeline@v0.9.0 · 5886 in / 1517 out tokens · 31000 ms · 2026-05-19T10:29:49.684451+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

53 extracted references · 53 canonical work pages · 6 internal anchors

  1. [1]

    Journal of Adolescent & Adult Literacy 56(2), 98–102 (2012)

    Jacobs, G.E.: The proverbial rock and hard place: The realities and risks of teach- ing in a world of multiliteracies, participatory culture, and mandates. Journal of Adolescent & Adult Literacy 56(2), 98–102 (2012)

    work page 2012

  2. [2]

    The Reading Teacher 75(4), 521–525 (2022)

    Opatz, M.O., Nelson, E.T.: Building bridges: Curating text sets to connect to learners’ lives. The Reading Teacher 75(4), 521–525 (2022)

    work page 2022

  3. [3]

    Journal of Adolescent & Adult Literacy 57(7), 565–574 (2014)

    Elish-Piper, L., Wold, L.S., Schwingendorf, K.: Scaffolding high school students’ reading of complex texts using linked text sets. Journal of Adolescent & Adult Literacy 57(7), 565–574 (2014)

    work page 2014

  4. [4]

    Journal of Adolescent & Adult Literacy 61(4), 433–444 (2018)

    Lupo, S.M., Strong, J.Z., Lewis, W., Walpole, S., McKenna, M.C.: Building back- ground knowledge through reading: Rethinking text sets. Journal of Adolescent & Adult Literacy 61(4), 433–444 (2018)

    work page 2018

  5. [5]

    The Reading Teacher 73(4), 513–524 (2020)

    Lupo, S.M., Berry, A., Thacker, E., Sawyer, A., Merritt, J.: Rethinking text sets to support knowledge building and interdisciplinary learning. The Reading Teacher 73(4), 513–524 (2020)

    work page 2020

  6. [6]

    : Gene ontology: tool for the unification of biology

    Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H., Cherry, J.M., Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T., et al. : Gene ontology: tool for the unification of biology. Nature genetics 25(1), 25–29 (2000)

    work page 2000

  7. [7]

    : Drug target ontology to classify and integrate drug discovery data

    Lin, Y., Mehta, S., K¨ u¸ c¨ uk-McGinty, H., Turner, J.P., Vidovic, D., Forlin, M., Koleti, A., Nguyen, D.-T., Jensen, L.J., Guha, R., et al. : Drug target ontology to classify and integrate drug discovery data. Journal of biomedical semantics 8, 1–16 (2017)

    work page 2017

  8. [8]

    In: Anatomy Ontologies for Bioinformatics: Principles and Practice, pp

    Haendel, M.A., Neuhaus, F., Osumi-Sutherland, D., Mabee, P.M., Mejino Jr, J.L., Mungall, C.J., Smith, B.: Caro–the common anatomy reference ontology. In: Anatomy Ontologies for Bioinformatics: Principles and Practice, pp. 327–349. Springer, London (2008)

    work page 2008

  9. [9]

    Nucleic acids research 49(D1), 1207–1217 (2021)

    K¨ ohler, S., Gargano, M., Matentzoglu, N., Carmody, L.C., Lewis-Smith, D., Vasilevsky, N.A., Danis, D., Balagura, G., Baynam, G., Brower, A.M., et al.: The 17 human phenotype ontology in 2021. Nucleic acids research 49(D1), 1207–1217 (2021)

    work page 2021

  10. [10]

    Mcginty, H.K.: Knowledge acquisition and representation methodology (knarm) and its applications. (2018). https://api.semanticscholar.org/CorpusID:64864929

    work page 2018

  11. [11]

    : Knowledge graphs

    Hogan, A., Blomqvist, E., Cochez, M., d’Amato, C., Melo, G.D., Gutierrez, C., Kirrane, S., Gayo, J.E.L., Navigli, R., Neumaier, S., et al. : Knowledge graphs. ACM Computing Surveys (Csur) 54(4), 1–37 (2021)

    work page 2021

  12. [12]

    In: 2022 7th IEEE International Conference on Data Science in Cyberspace (DSC), pp

    Yan, Q., Fan, J., Li, M., Qu, G., Xiao, Y.: A survey on knowledge graph embed- ding. In: 2022 7th IEEE International Conference on Data Science in Cyberspace (DSC), pp. 576–583 (2022). IEEE

    work page 2022

  13. [13]

    Symmetry 13(3), 485 (2021)

    Wang, M., Qiu, L., Wang, X.: A survey on knowledge graph embeddings for link prediction. Symmetry 13(3), 485 (2021)

    work page 2021

  14. [14]

    Machine Learning 110(7), 1813–1845 (2021)

    Chen, J., Hu, P., Jimenez-Ruiz, E., Holter, O.M., Antonyrajah, D., Horrocks, I.: Owl2vec*: Embedding of owl ontologies. Machine Learning 110(7), 1813–1845 (2021)

    work page 2021

  15. [15]

    In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp

    Perozzi, B., Al-Rfou, R., Skiena, S.: Deepwalk: Online learning of social represen- tations. In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 701–710 (2014)

    work page 2014

  16. [16]

    Efficient Estimation of Word Representations in Vector Space

    Mikolov, T.: Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781 3781 (2013)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  17. [17]

    In: Recommender Systems Handbook, pp

    Ricci, F., Rokach, L., Shapira, B.: Introduction to recommender systems hand- book. In: Recommender Systems Handbook, pp. 1–35. Springer, ??? (2010)

    work page 2010

  18. [18]

    International Journal of Computer Applications 110(4) (2015)

    Thorat, P.B., Goudar, R.M., Barve, S.: Survey on collaborative filtering, content- based filtering and hybrid recommendation system. International Journal of Computer Applications 110(4) (2015)

    work page 2015

  19. [19]

    Computing 100, 809–823 (2018)

    Yang, Q.: A novel recommendation system based on semantics and context awareness. Computing 100, 809–823 (2018)

    work page 2018

  20. [20]

    The Reading Teacher 73(5), 575–586 (2020)

    Robertson, D.A., Padesky, C.J.: Keeping students interested: Interest-based instruction as a tool to engage. The Reading Teacher 73(5), 575–586 (2020)

    work page 2020

  21. [21]

    TEFLIN Journal: A Publication on the Teaching & Learning of English 32(2) (2021)

    Tan, D.N.L., Mante-Estacio, M.J.: Reader-text connection: Reporting the engage- ment of high school students with culturally relevant texts. TEFLIN Journal: A Publication on the Teaching & Learning of English 32(2) (2021)

    work page 2021

  22. [22]

    Literacy Practice and Research 47(3), 3 (2022) 18

    Graff, J., Liang, L.A., Martinez, M.G., McClure, A., Day, D., Sableski, M.-K., Arnold, J.M.: Contemporary children’s literature in education courses: Diverse, complex, and critical. Literacy Practice and Research 47(3), 3 (2022) 18

    work page 2022

  23. [23]

    The Reading Teacher 70(5), 629–633 (2017)

    Yokota, J., Teale, W.H.: Striving for international understanding through litera- ture. The Reading Teacher 70(5), 629–633 (2017)

    work page 2017

  24. [24]

    Kansas English 105 (2024)

    Rios, A.: The kansas literary canon: A study of texts taught in high school english courses. Kansas English 105 (2024)

    work page 2024

  25. [25]

    English Education 47(3), 245–275 (2015)

    Watkins, N., Ostenson, J.: Navigating the text selection gauntlet: Exploring fac- tors that influence english teachers’ choices. English Education 47(3), 245–275 (2015)

    work page 2015

  26. [26]

    Journal of Adolescent & Adult Literacy 65(4), 321–329 (2022)

    Ervin, J.: Critically reading the canon: Culturally sustaining approaches to a prescribed literature curriculum. Journal of Adolescent & Adult Literacy 65(4), 321–329 (2022)

    work page 2022

  27. [27]

    ALAN Review 48(2) (2021)

    Glaws, A.L.: Conversations with classroom teachers: Surveying teachers’ use and perceptions of middle grades and young adult literature. ALAN Review 48(2) (2021)

    work page 2021

  28. [28]

    just plain reading

    Ivey, G., Broaddus, K.: “just plain reading”: A survey of what makes students want to read in middle school classrooms. Reading research quarterly 36(4), 350– 377 (2001)

    work page 2001

  29. [29]

    Bioinformatics34(13), 52–60 (2018)

    Smaili, F.Z., Gao, X., Hoehndorf, R.: Onto2vec: joint vector-based representation of biological entities and their ontology-based annotations. Bioinformatics34(13), 52–60 (2018)

    work page 2018

  30. [30]

    Bioinformatics 35(12), 2133–2140 (2019)

    Smaili, F.Z., Gao, X., Hoehndorf, R.: Opa2vec: combining formal and infor- mal content of biomedical ontologies to improve similarity-based prediction. Bioinformatics 35(12), 2133–2140 (2019)

    work page 2019

  31. [31]

    APSIPA Transactions on Signal and Information Processing 9, 15 (2020)

    Chen, F., Wang, Y.-C., Wang, B., Kuo, C.-C.J.: Graph representation learning: a survey. APSIPA Transactions on Signal and Information Processing 9, 15 (2020)

    work page 2020

  32. [32]

    Briefings in Bioinformatics 23(1), 340 (2022)

    Yi, H.-C., You, Z.-H., Huang, D.-S., Kwoh, C.K.: Graph representation learning in bioinformatics: trends, methods and applications. Briefings in Bioinformatics 23(1), 340 (2022)

    work page 2022

  33. [33]

    ACM Transactions on Intelligent Systems and Technology 15(1), 1–55 (2024)

    Khoshraftar, S., An, A.: A survey on graph representation learning methods. ACM Transactions on Intelligent Systems and Technology 15(1), 1–55 (2024)

    work page 2024

  34. [34]

    Physica A: Statistical Mechanics and its Applications 553, 124289 (2020)

    Kumar, A., Singh, S.S., Singh, K., Biswas, B.: Link prediction techniques, appli- cations, and performance: A survey. Physica A: Statistical Mechanics and its Applications 553, 124289 (2020)

    work page 2020

  35. [35]

    Social network data analytics, 115–148 (2011)

    Bhagat, S., Cormode, G., Muthukrishnan, S.: Node classification in social networks. Social network data analytics, 115–148 (2011)

    work page 2011

  36. [36]

    ACM Transactions on Knowledge Discovery from Data (TKDD) 15(2), 1–49 (2021)

    Rossi, A., Barbosa, D., Firmani, D., Matinata, A., Merialdo, P.: Knowledge graph 19 embedding for link prediction: A comparative analysis. ACM Transactions on Knowledge Discovery from Data (TKDD) 15(2), 1–49 (2021)

    work page 2021

  37. [37]

    IEEE Transactions on Knowledge and Data Engineering 34(8), 3549–3568 (2020)

    Guo, Q., Zhuang, F., Qin, C., Zhu, H., Xie, X., Xiong, H., He, Q.: A survey on knowledge graph-based recommender systems. IEEE Transactions on Knowledge and Data Engineering 34(8), 3549–3568 (2020)

    work page 2020

  38. [38]

    In: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, pp

    Yang, Y., Huang, C., Xia, L., Li, C.: Knowledge graph contrastive learning for rec- ommendation. In: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 1434–1443 (2022)

    work page 2022

  39. [39]

    Fast and scalable learning of neuro-symbolic representations of biomedical knowledge

    Agibetov, A., Samwald, M.: Fast and scalable learning of neuro-symbolic repre- sentations of biomedical knowledge. arXiv preprint arXiv:1804.11105 (2018)

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  40. [40]

    Bioinformatics 33(17), 2723–2730 (2017)

    Alshahrani, M., Khan, M.A., Maddouri, O., Kinjo, A.R., Queralt-Rosinach, N., Hoehndorf, R.: Neuro-symbolic representation learning on biological knowledge graphs. Bioinformatics 33(17), 2723–2730 (2017)

    work page 2017

  41. [41]

    In: CEUR Workshop Proceedings, vol

    Rosati, J., Ristoski, P., Di Noia, T., Leone, R.d., Paulheim, H.: Rdf graph embed- dings for content-based recommender systems. In: CEUR Workshop Proceedings, vol. 1673, pp. 23–30 (2016). RWTH Aachen

    work page 2016

  42. [42]

    In: Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, pp

    Cochez, M., Ristoski, P., Ponzetto, S.P., Paulheim, H.: Biased graph walks for rdf graph embeddings. In: Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, pp. 1–12 (2017)

    work page 2017

  43. [43]

    In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp

    Grover, A., Leskovec, J.: node2vec: Scalable feature learning for networks. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 855–864 (2016)

    work page 2016

  44. [44]

    In: Companion Proceedings of the ACM Web Conference 2024, pp

    De Nadai, M., Fabbri, F., Gigioli, P., Wang, A., Li, A., Silvestri, F., Kim, L., Lin, S., Radosavljevic, V., Ghael, S., et al.: Personalized audiobook recommendations at spotify through graph neural networks. In: Companion Proceedings of the ACM Web Conference 2024, pp. 403–412 (2024)

    work page 2024

  45. [45]

    Semi-Supervised Classification with Graph Convolutional Networks

    Kipf, T.N., Welling, M.: Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 (2016)

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  46. [46]

    Graph Attention Networks

    Veliˇ ckovi´ c, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., Bengio, Y.: Graph attention networks. arXiv preprint arXiv:1710.10903 (2017)

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  47. [47]

    https://www.nysed.gov/ sites/default/files/programs/standards-instruction/appendix-a-common-core-s tandards-for-ela-literacy-supporting-research-glossary.pdf

    New York State Education Department: Appendix A: Common Core Standards for ELA/Literacy – Supporting Research and Glossary. https://www.nysed.gov/ sites/default/files/programs/standards-instruction/appendix-a-common-core-s tandards-for-ela-literacy-supporting-research-glossary.pdf. Accessed: 2025-05-05 (n.d.) 20

    work page 2025

  48. [48]

    American Secondary Education, 31–45 (2016)

    Rybakova, K., Roccanti, R.: Connecting the canon to current young adult literature. American Secondary Education, 31–45 (2016)

    work page 2016

  49. [49]

    In: International Conference on Knowledge Engineering and Ontology Development, vol

    Raad, J., Cruz, C.: A survey on ontology evaluation methods. In: International Conference on Knowledge Engineering and Ontology Development, vol. 2, pp. 179–186 (2015). SciTePress

    work page 2015

  50. [50]

    Embedding Entities and Relations for Learning and Inference in Knowledge Bases

    Yang, B., Yih, W.-t., He, X., Gao, J., Deng, L.: Embedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575 (2014)

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  51. [51]

    Advances in neural information processing systems 26 (2013)

    Bordes, A., Usunier, N., Garcia-Duran, A., Weston, J., Yakhnenko, O.: Trans- lating embeddings for modeling multi-relational data. Advances in neural information processing systems 26 (2013)

    work page 2013

  52. [52]

    https://arxiv.org/abs/2402.19195

    Madushanka, T., Ichise, R.: Negative Sampling in Knowledge Graph Representa- tion Learning: A Review (2024). https://arxiv.org/abs/2402.19195

    work page arXiv 2024

  53. [53]

    From Local to Global: A Graph RAG Approach to Query-Focused Summarization

    Edge, D., Trinh, H., Cheng, N., Bradley, J., Chao, A., Mody, A., Truitt, S., Metropolitansky, D., Ness, R.O., Larson, J.: From local to global: A graph rag approach to query-focused summarization. arXiv preprint arXiv:2404.16130 (2024) 21

    work page internal anchor Pith review Pith/arXiv arXiv 2024