Elamur Research⌕
☰
home / papers / 2020 / Joint Reasoning for Multi-Faceted Commonsense Knowledge

Joint Reasoning for Multi-Faceted Commonsense Knowledge

Paper · 2020 · Yohan Chalier, Simon Razniewski and Gerhard Weikum

Summary

Existing commonsense knowledge (CSK) resources such as ConceptNet store concept–property statements with a single confidence score. They treat each concept separately, and they cannot tell apart a property that is merely true, one that is common, one that is distinctive, and one that people would name first. The authors propose Dice (Diverse Commonsense Knowledge), which scores every statement along four facets: plausibility, typicality, remarkability and saliency. Related concepts are found through taxonomy neighbourhoods (siblings, parents) or word-embedding similarity. Dice first estimates prior scores from cues via regression. It then reasons jointly using weighted soft constraints that link the facets across related concepts. The problem is encoded as an integer linear program, and the reduced costs of its LP relaxation are used to produce graded rankings for each facet. Dice was applied to ConceptNet, TupleKB and Quasimodo and evaluated against human judgements, where it outperformed single-score baselines. The result is cleaner, richer CSK for applications such as dialogue systems and image description.

Key points

  • The four facets are plausibility, typicality, remarkability and saliency. For example, hyenas eating carcasses is salient, while hyenas drinking water is typical but unremarkable.
  • The enriched knowledge bases contain over 1.6M statements about roughly 74k concepts. Six datasets were released: three scored collections plus inferred statements that add about 50% more data. A demo is available at dice.mpi-inf.mpg.de.
  • In the ablation, the full Dice method reached an average ppref of 0.66, compared with 0.58 for priors alone and 0.51 for reasoning alone. Across facets, Dice scored 0.58–0.69 ppref against human rankings, and plausibility was the hardest facet.
  • Hand-built taxonomies improved ppref by 0.03–0.11 compared with the noisy, web-derived WebIsALOD taxonomy.
  • Of the three inputs, the crowdsourced ConceptNet was the cleanest and easiest to process; Quasimodo and TupleKB were noisier.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🔬Max Planck SocietyMunich · Germany19 works · 42,336 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 19,503 worksArtificial Intelligence · 18,805 worksArts and Humanities · 17,959 worksLanguage and Linguistics · 17,789 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 1,948 works

Cite

Cite as Yohan Chalier, Simon Razniewski and Gerhard Weikum. 2020. Joint Reasoning for Multi-Faceted Commonsense Knowledge. arXiv:2001.04170 (AKBC 2020).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2001.04170 · arXiv 2001.04170

@misc{chalier2020joint,
  title = {{Joint Reasoning for Multi-Faceted Commonsense Knowledge}},
  author = {Yohan Chalier, Simon Razniewski and Gerhard Weikum},
  year = {2020},
  doi = {10.48550/arxiv.2001.04170},
  eprint = {2001.04170},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2001.04170}
}

References

49 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Tamas Abraham and John F Roddick. Survey of spatio-temporal databases. GeoInformatica, 1999. find
  2. Aishwarya Agrawal, Jiasen Lu, Stanislaw Antol, Margaret Mitchell, C. Lawrence Zitnick, Devi Parikh, and Dhruv Batra. VQA: visual question answering. IJCV, 2017. find
  3. Sören Auer, Christian Bizer, Georgi Kobilarov, Jens Lehmann, Richard Cyganiak, and Zachary G. Ives. DBpedia: A nucleus for a web of open data. ISWC, 2007. find
  4. James Bergstra, Dan Yamins, and David D. Cox. Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures. ICML, 2013. find
  5. Dimitris Bertsimas and John N Tsitsiklis. Introduction to linear optimization. Athena Scientific, 1997. find
  6. Antoine Bosselut, Hannah Rashkin, Maarten Sap, Chaitanya Malaviya, Asli Çelikyilmaz, and Yejin Choi. COMET: commonsense transformers for automatic knowledge graph construction. ACL, 2019.
  7. Samuel R. Bowman, Gabor Angeli, Christopher Potts, and Christopher D. Manning. A large annotated corpus for learning natural language inference. EMNLP, 2015.
  8. Andrew Carlson, Justin Betteridge, Bryan Kisiel, Burr Settles, Estevam R Hruschka, and Tom M Mitchell. Toward an architecture for never-ending language learning. AAAI, 2010. find
  9. Ben Carterette, Paul N Bennett, David Maxwell Chickering, and Susan T Dumais. Here or there: Preference judgments for relevance. ECIR, 2008. find
  10. Xuelu Chen, Muhao Chen, Weijia Shi, Yizhou Sun, and Carlo Zaniolo. Embedding uncertain knowledge graphs. AAAI, 2019. find
  11. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding. NAACL, 2019.
  12. Pedro M. Domingos and Daniel Lowd. Markov Logic: An Interface Layer for Artificial Intelligence. Morgan & Claypool, 2009. find
  13. Matt Gardner, Joel Grus, Mark Neumann, Oyvind Tafjord, Pradeep Dasigi, Nelson F. Liu, Matthew Peters, Michael Schmitz, and Luke Zettlemoyer. AllenNLP: A deep semantic natural language processing platform. Workshop for NLP Open Source Software (NLP-OSS), 2018.
  14. LLC Gurobi Optimization. Gurobi optimizer reference manual, 2019. URL http://www.gurobi.com. ↗
  15. Sebastian Hellmann, Claus Stadler, Jens Lehmann, and Sören Auer. Dbpedia live extraction. OTM, 2009. find
  16. Sven Hertling and Heiko Paulheim. Webisalod: Providing hypernymy relations extracted from the web as linked open data. ISWC, 2017. find
  17. Sven Hertling and Heiko Paulheim. Webisalod: providing hypernymy relations extracted from the web as linked open data. International Semantic Web Conference, pages 111–119. Springer, 2017. find
  18. Johannes Hoffart, Fabian M. Suchanek, Klaus Berberich, and Gerhard Weikum. Yago2: A spatially and temporally enhanced knowledge base from wikipedia. Artif. Intell. 194: 28-61, 2013. find
  19. Andrej Karpathy and Li Fei-Fei. Deep visual-semantic alignments for generating image descriptions. Trans. Pattern Anal. Mach. Intell., 2017. find
  20. Douglas B Lenat. Cyc: A large-scale investment in knowledge infrastructure. Communications of the ACM, 1995. find
  21. Vasco Manquinho, Joao Marques-Silva, and Jordi Planes. Algorithms for weighted boolean optimization. SAT, 2009. find
  22. Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. Distributed representations of words and phrases and their compositionality. NIPS, 2013.
  23. George A. Miller. Wordnet: A lexical database for English. CACM, 1995.
  24. Bhavana Dalvi Mishra, Niket Tandon, and Peter Clark. Domain-targeted, high precision knowledge extraction. TACL, 2017.
  25. Natalya Fridman Noy, Yuqing Gao, Anshu Jain, Anant Narayanan, Alan Patterson, and Jamie Taylor. Industry-scale knowledge graphs: lessons and challenges. Communications of the ACM, 2019. find
  26. Ana Ozaki, Markus Krötzsch, and Sebastian Rudolph. Happy ever after: Temporally attributed description logics. Description Logics, 2018. find
  27. Peter F Patel-Schneider. Contextualization via qualifiers. Workshop on Contextualized Knowledge Graphs, 2018. find
  28. Jeffrey Pennington, Richard Socher, and Christopher Manning. Glove: Global vectors for word representation. EMNLP, 2014.
  29. Jay Pujara, Hui Miao, Lise Getoor, and William Cohen. Knowledge graph identification. ISWC, 2013. find
  30. Prabhakar Raghavan and Clark D. Thompson. Randomized rounding: a technique for provably good algorithms and algorithmic proofs. Combinatorica, 1987. find
  31. Nazneen Fatema Rajani, Bryan McCann, Caiming Xiong, and Richard Socher. Explain yourself! leveraging language models for commonsense reasoning. ACL, 2019.
  32. Julien Romero, Simon Razniewski, Koninika Pal, Jeff Z. Pan, Archit Sakhadeo, and Gerhard Weikum. Commonsense properties from query logs and question answering forums. CIKM, 2019. find
  33. Maarten Sap, Ronan LeBras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A Smith, and Yejin Choi. Atomic: An atlas of machine commonsense for if-then reasoning. AAAI, 2018. find
  34. Carissa Schoenick, Peter Clark, Oyvind Tafjord, Peter D. Turney, and Oren Etzioni. Moving beyond the Turing test with the Allen AI science challenge. Commun. ACM, 2017.
  35. Jaeho Shin, Sen Wu, Feiran Wang, Christopher De Sa, Ce Zhang, and Christopher Ré. Incremental knowledge base construction using deepdive. VLDB, 2015. find
  36. Kurt Shuster, Samuel Humeau, Hexiang Hu, Antoine Bordes, and Jason Weston. Engaging image captioning via personality. CVPR, 2019. find
  37. Robyn Speer and Catherine Havasi. ConceptNet 5: A large semantic network for relational knowledge. Theory and Applications of Natural Language Processing, 2012. find
  38. Vivek Srikumar and Dan Roth. A joint model for extended semantic role labeling. EMNLP, 2011. find
  39. Fabian M Suchanek, Gjergji Kasneci, and Gerhard Weikum. Yago: a core of semantic knowledge. WWW, 2007. find
  40. Fabian M Suchanek, Mauro Sozio, and Gerhard Weikum. Sofie: a self-organizing framework for information extraction. WWW, 2009. find
  41. Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. CommonsenseQA: A question answering challenge targeting commonsense knowledge. NAACL, 2019.
  42. Niket Tandon, Gerard de Melo, Fabian M. Suchanek, and Gerhard Weikum. WebChild: harvesting and organizing commonsense knowledge from the web. WSDM, 2014. find
  43. Niket Tandon, Gerard de Melo, and Gerhard Weikum. WebChild 2.0 : Fine-grained commonsense knowledge distillation. ACL, 2017. find
  44. Vijay V Vazirani. Approximation algorithms. Springer Science & Business Media, 2013. find
  45. Denny Vrandečić and Markus Krötzsch. Wikidata: a free collaborative knowledgebase. CACM, 2014.
  46. Michael L. Wick, Andrew McCallum, and Gerome Miklau. Scalable probabilistic databases with factor graphs and MCMC. PVLDB, 2010. find
  47. Mohamed Yahya, Denilson Barbosa, Klaus Berberich, Qiuyue Wang, and Gerhard Weikum. Relationship queries on extended knowledge graphs. WSDM, 2016. find
  48. Pengcheng Yang, Lei Li, Fuli Luo, Tianyu Liu, and Xu Sun. Enhancing topic-to-essay generation with external commonsense knowledge. ACL, 2019. find
  49. Tom Young, Erik Cambria, Iti Chaturvedi, Hao Zhou, Subham Biswas, and Minlie Huang. Augmenting end-to-end dialogue systems with commonsense knowledge. AAI, 2018. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/chalier-2020-joint-reasoning-multi-faceted-commonsense/