Source Citations & Attribution

This project stands on the shoulders of giants. I wish to express my profound gratitude and admiration for the pioneering researchers whose work provided the foundational pillars for this endeavor. The GDS architecture is a synthesis of their brilliant ideas, and this page serves to formally acknowledge their contributions.

This annex consolidates the formal citations, acknowledgements, and licensing notes for every external resource that powers the GDS lexicon builder and downstream runtime. Please reference these works in any public release, publication, or demo derived from this project.

Foundational Concepts & Algorithms

This section acknowledges the core scientific and algorithmic concepts that inspired and enabled the GDS architecture.

Hyperdimensional Computing (HDC)

  • Citation: Kanerva, P. (2009). Hyperdimensional computing: An introduction to computing in distributed representation with high-dimensional random vectors. Cognitive Computation, 1(2), 139-159.
  • Notes: This work provides the mathematical and conceptual foundation for using high-dimensional binary vectors, which is central to GDS’s SemanticParticle representation and its compositional operators.

The Free Energy Principle

  • Citation: Friston, K. (2010). The free-energy principle: a unified brain theory? Nature Reviews Neuroscience, 11(2), 127-138.
  • Notes: Friston’s principle provides the neuroscientific inspiration for GDS’s learning paradigm, where the system acts to minimize surprise by modifying the geometry of its conceptual space.

General Relativity as a Semantic Analogy

  • Honorary Mention: Einstein, A. (1916). Die Grundlage der allgemeinen Relativitätstheorie (The Foundation of the General Theory of Relativity). Annalen der Physik, 354(7), 769-822.
  • Notes: While not a direct implementation, Einstein’s theory provides the core conceptual metaphor for GDS: concepts as mass warping a semantic spacetime, and reasoning as movement along geodesics.

K-Shortest Path Algorithm

  • Citation: Yen, J. Y. (1971). Finding the K Shortest Loopless Paths in a Network. Management Science, 17(11), 712-716.
  • Notes: Yen’s algorithm is implemented in the Reasoner to find not just the optimal path, but also the runner-up (“second-best”) path, which is critical for the margin-based contrastive learning loop.

Knowledge Graphs & Embeddings

BabelNet

  • Citation: R. Navigli, M. Bevilacqua, S. Conia, D. Montagnini, F. Cecconi. Ten Years of BabelNet: A Survey. Proceedings of IJCAI 2021, pp. 4559–4567.
  • Notes: Required for all usages of the daily enrichment pipeline. Access granted under a non-commercial research agreement; respect BabelNet’s license.

ConceptNet 5.x

  • Citation: Robyn Speer, Joshua Chin, Catherine Havasi. ConceptNet 5.5: An Open Multilingual Graph of General Knowledge. AAAI 2017, pp. 4444–4451. http://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14972
  • Notes: Dataset distributed under CC BY-SA 4.0. The same reference covers both the graph (assertions.csv) and the downstream integration in the Reasoner.

ConceptNet Numberbatch

  • Citation (recommended by maintainers): Robyn Speer, Joshua Chin, Catherine Havasi. ConceptNet 5.5: An Open Multilingual Graph of General Knowledge. AAAI 2017, pp. 4444–4451.
  • Supplementary attributions (per README): The embeddings ensemble incorporates GloVe, word2vec, OpenSubtitles 2016, and fastText. Cite these works if their contributions are highlighted separately.
  • Notes: Embeddings fetched from https://github.com/commonsense/conceptnet-numberbatch (CC BY-SA 4.0).

Lexicons & WordNets

Open English WordNet (EN)

  • Citation: John P. McCrae, Alexandre Rademaker, Francis Bond, Ewa Rudnicka, Christiane Fellbaum. English WordNet 2019 – An Open-Source WordNet for English. Proceedings of the 10th Global Wordnet Conference, 2019.
  • Notes: Licensed under CC BY 4.0. Project site: https://en-word.net/.

OdeNet – Open German WordNet (DE)

  • Citation: Melanie Siegel, Francis Bond. OdeNet: Compiling a German Wordnet from Other Resources. Proceedings of the 11th Global Wordnet Conference (GWC 2021), pp. 192–198. https://www.aclweb.org/anthology/2021.gwc-1.22
  • Notes: Data under CC BY-SA 4.0. Original README distributed with the dump: data/raw/german_wordnet/README.md.

Romanian WordNet (RoWordNet via RoLLOD)

  • Primary citation: Dan Tufiș, Verginica Barbu Mititelu. The Lexical Ontology for Romanian. In Language Production, Cognition, and the Lexicon, Text, Speech and Language Technology, vol. 48, Springer, 2014, pp. 491–504.
  • API/tooling citation (optional): Ștefan Daniel Dumitrescu, Andrei Marius Avram, Luciana Morogan, Ștefan-Adrian Toma. RoWordNet – A Python API for the Romanian WordNet. ECAI 2018.
  • Notes: Distributed under CC BY-SA 4.0; see data/oewm_lexicons/LICENSE.json.

SEQUOIA Trilingual Lexicon Builder (OEWM Integration)

  • Citation: Internal toolchain authored by Mihai Adrian Mateescu (Profit Minds). When publishing combined lexicon outputs, attribute the upstream sources listed above and note the CC BY-SA 4.0 redistribution requirement (data/oewm_lexicons/README.md).

Affective & Emotion Resources

NRC VAD Lexicon v2

  • Primary citation: Saif M. Mohammad. NRC VAD Lexicon v2: Norms for Valence, Arousal, and Dominance for over 55k English Terms. arXiv:2503.23547, 2025.
  • Foundational citation (per README): Saif M. Mohammad. Obtaining Reliable Human Ratings of Valence, Arousal, and Dominance for 20,000 English Words. Proceedings of ACL 2018.
  • Notes: Point to the homepage http://saifmohammad.com/WebPages/nrc-vad.html in documentation. Usage restricted to research; retain copyright notice.

German Affective Norms (de_emo_norms.txt)

  • Citation: Maximilian Köper, Sabine Schulte im Walde. Analogies in Complex Verb Meaning Shifts: The Effect of Affect in Semantic Similarity Models. NAACL 2018.
  • Additional acknowledgement (per dataset page): Cite the training resources underpinning the automatically generated norms when relevant (NRC Hashtag Emotion Lexicon, Warriner et al. 2013, Dodds et al. 2011, Brysbaert et al. 2014).
  • Notes: Dataset obtained from the IMS Stuttgart resource portal: https://www.ims.uni-stuttgart.de/en/research/resources/experiment-data/de-affect-norms/.

Operational Requirements

  • Maintain the license headers already present in source and data files.
  • Include these citations in README files, academic papers, slide decks, and any GitHub Pages deployments derived from the Quarto site.
  • For commercial inquiries, contact the respective data providers; most resources listed here are available only for research/non-commercial use (BabelNet, NRC VAD, IMS emotion norms).
"