IP Library › Granted Patent US 12,307,382
Granted Patent B1
US 12,307,382 · App. 16/592,115 · Granted May 20, 2025

Neural taxonomy expander

Inventors: Emaad Ahmed Manzoor (Pittsburgh, PA); Rui Li (San Jose, CA); Dhananjay Shrouty (San Francisco, CA); Jurij Leskovec (Stanford, CA)
Assignee: Pinterest, Inc.
G06N5/022G06N3/04G06N3/08
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,307,382
App. No.
16/592,115
Granted
May 20, 2025
Kind
B1
Abstract

Systems and methods for automatically placing a taxonomy candidate within an existing taxonomy are presented. More particularly, a neural taxonomy expander (a neural network model) is trained according to the existing, curated taxonomic hierarchy. Moreover, for each node in the taxonomic hierarchy, an embedding vector is generated. A taxonomy candidate is received, where the candidate is to be placed within the existing taxonomy. An embedding vector is generated for the candidate and projected by a projection function of the neural taxonomy expander into the taxonomic hyperspace. A set of closest neighbors to the projected embedding vector of the taxonomy candidate is identified and the closest neighbor of the set is assumed as the parent for the taxonomy candidate. The taxonomy candidate is added to the existing taxonomic hierarchy as a child to the identified parent node.

Claims (76)

1. A computer-implemented method for automatically locating a taxonomy candidate into an existing taxonomic hierarchy, the computer-implemented method comprising:

training a neural taxonomy expander according to the existing taxonomic hierarchy, wherein training the neural taxonomy expander includes:

generating a plurality of candidate/hypernym pairs, the plurality of candidate/hypernym pairs including positive candidate/hypernym pairs and negative candidate/hypernym pairs;

subdividing the plurality of candidate/hypernym pairs into a training set and a validation set;

repeatedly, until a predetermined loss threshold from processing the validation set is met:

training the neural taxonomy expander according to the training set until the predetermined loss threshold is met in processing the training set;

validating the neural taxonomy expander according to the validation set; and

updating the neural taxonomy expander according to results of the training upon a determination that the predetermined loss threshold for the validation set is not met; and

generating an executable neural taxonomy expander according to the trained neural taxonomy expander;

generating an embedding vector of the taxonomy candidate;

projecting the embedding vector of the taxonomy candidate into a taxonomic hyperspace according to a taxonomy projection by the executable neural taxonomy expander;

identifying a set of closest neighbors of the existing taxonomic hierarchy to the projected embedding vector of the taxonomy candidate;

determining a closest neighbor of the identified set of closest neighbors as an immediate parent of the taxonomy candidate; and

adding the taxonomy candidate into the existing taxonomic hierarchy as a child node of the determined immediate parent.

2. The computer-implemented method of claim 1 , wherein generating the embedding vector of the taxonomy candidate comprises:

aggregating textual content relating to the taxonomy candidate; and

generating the embedding vector of the taxonomy candidate according to the aggregated textual content relating to the taxonomy candidate.

3. The computer-implemented method of claim 2 , wherein the aggregated textual content relating to the taxonomy candidate comprises any one or more of:

a title associated with the taxonomy candidate;

one or more captions associated with the taxonomy candidate;

collection titles of collections in which the taxonomy candidate is a member;

a URL (uniform resource locator) or URI (uniform resource identifier) associated with the taxonomy candidate;

one or more user comments made in regard to the taxonomy candidate; and

textual content descriptive of a subject matter of the taxonomy candidate.

4. The computer-implemented method of claim 2 , further comprising projecting each of a plurality of nodes of the existing taxonomic hierarchy into the taxonomic hyperspace according to the taxonomy projection of the executable neural taxonomy expander.

5. The computer-implemented method of claim 4 , wherein projecting each of the plurality of nodes of the existing taxonomic hierarchy into the taxonomic hyperspace according to the taxonomy projection of the executable neural taxonomy expander comprises generating a respective embedding vector for each of the plurality of nodes of the existing taxonomic hierarchy.

6. The computer-implemented method of claim 5 , wherein generating a respective embedding vector for each of the plurality of nodes of the existing taxonomic hierarchy comprises, for each of the plurality of nodes of the existing taxonomic hierarchy:

aggregating second textual content relating to a current node of the existing taxonomic hierarchy; and

generating the respective embedding vector of the current node of the existing taxonomic hierarchy according to the aggregated second textual content relating to the current node of the existing taxonomic hierarchy.

7. The computer-implemented method of claim 1 , wherein:

each positive candidate/hypernym pair includes a positive child node and an immediate positive parent node; and

each negative candidate/hypernym pair includes a negative child node and a hypernym that is not an immediate parent node of the negative child node.

8. The computer-implemented method of claim 7 , wherein generating the plurality of candidate/hypernym pairs from the existing taxonomic hierarchy comprises generating a first positive candidate/hypernym pair for each child node in the existing taxonomic hierarchy.

9. The computer-implemented method of claim 8 , wherein generating the plurality of candidate/hypernym pairs from the existing taxonomic hierarchy comprises generating a plurality of negative candidate/hypernym pairs for each parent node in the existing taxonomic hierarchy proportional to a number of immediate children of the parent node.

10. A computer-readable medium bearing computer executable instructions which, when executed on a computing system, carry out a method for automatically locating a taxonomy candidate into a taxonomic hierarchy, the method comprising:

training a neural taxonomy expander according to the taxonomic hierarchy, wherein training the neural taxonomy expander includes:

generating a plurality of candidate/hypernym pairs, the plurality of candidate/hypernym pairs including positive candidate/hypernym pairs and negative candidate/hypernym pairs;

subdividing the plurality of candidate/hypernym pairs into a training set and a validation set;

repeatedly, until a predetermined loss threshold from processing the validation set is met:

training the neural taxonomy expander according to the training set until the predetermined loss threshold is met in processing the training set;

validating the neural taxonomy expander according to the validation set; and

updating the neural taxonomy expander according to results of the training upon a determination that the predetermined loss threshold for the validation set is not met; and

generating an executable neural taxonomy expander according to the trained neural taxonomy expander;

generating, for each node in the taxonomic hierarchy, a respective embedding vector;

generating a candidate embedding vector of the taxonomy candidate;

projecting the candidate embedding vector of the taxonomy candidate into a taxonomic hyperspace according to the taxonomy projection of the neural taxonomy expander;

identifying a set of closest neighbors in the taxonomic hierarchy to the projected candidate embedding vector of the taxonomy candidate;

determining a closest neighbor of the identified set of closest neighbors as an immediate parent of the taxonomy candidate; and

adding the taxonomy candidate into the taxonomic hierarchy as a child node of the determined immediate parent.

11. The computer-readable medium of claim 10 , wherein generating the candidate embedding vector of the taxonomy candidate comprises:

aggregating textual content relating to the taxonomy candidate; and

generating the candidate embedding vector of the taxonomy candidate according to the aggregated textual content relating to the taxonomy candidate.

12. The computer-readable medium of claim 11 , wherein:

each positive candidate/hypernym pair includes a positive child node and an immediate positive parent node; and

each negative candidate/hypernym pair includes a negative child node and a hypernym that is not an immediate parent node of the negative child node.

13. The computer-readable medium of claim 12 , wherein generating the plurality of candidate/hypernym pairs from the taxonomic hierarchy comprises generating a first positive candidate/hypernym pair for each child node in the taxonomic hierarchy.

14. The computer-readable medium of claim 13 , wherein generating the plurality of candidate/hypernym pairs from the taxonomic hierarchy comprises generating a plurality of negative candidate/hypernym pairs for each parent node in the taxonomic hierarchy proportional to a number of immediate children of the parent node.

15. The computer-readable medium of claim 14 , wherein generating the embedding vector for each node of the taxonomic hierarchy comprises:

aggregating textual content relating to the node of the taxonomic hierarchy; and

generating the embedding vector of the node of the taxonomic hierarchy according to the aggregated textual content relating to the node of the taxonomic hierarchy.

16. A computer system configured to train a neural taxonomy expander to automatically place a taxonomy candidate into an existing taxonomic hierarchy, the computer system comprising a processor and a memory, wherein the processor executes instructions stored in the memory such that execution of the instructions cause the computer system to:

generate a plurality of candidate/hypernym pairs from a taxonomic hierarchy, the plurality of candidate/hypernym pairs including positive candidate/hypernym pairs and negative candidate/hypernym pairs, wherein each positive candidate/hypernym pair includes a positive child node and an immediate positive parent node, and where each negative candidate/hypernym pair includes a negative child node and a hypernym that is not an immediate parent node of the negative child node;

subdivide the plurality of candidate/hypernym pairs into a training set and a validation set;

repeatedly, until a predetermined loss threshold from processing first candidate/hypernym pairs of the validation set is met:

train the neural taxonomy expander according to second candidate/hypernym pairs of the training set until the predetermined loss threshold is met in processing the training set;

validate the neural taxonomy expander according to the first candidate/hypernym pairs in the validation set; and

update the neural taxonomy expander according to results of the training upon a determination that the predetermined loss threshold from processing the first candidate/hypernym pairs of the validation set is not met; and

generate an executable neural taxonomy expander according to the trained neural taxonomy expander.

17. The computer system of claim 16 , wherein in generating the plurality of candidate/hypernym pairs from the taxonomic hierarchy, the computer system generates a first positive candidate/hypernym pair for each child node in the taxonomic hierarchy.

18. The computer system of claim 17 , wherein in generating the plurality of candidate/hypernym pairs from the taxonomic hierarchy, the computer system generates a plurality of negative candidate/hypernym pairs for each parent node in the taxonomic hierarchy proportional to a number of immediate children of the parent node.

19. The computer system of claim 18 , wherein in determining whether the predetermined loss threshold is met, the computing system is further configured to:

determine a Mean Reciprocal Rank (MRR) for at least some of the plurality of candidate/hypernym pairs; and

determine whether the predetermined loss threshold is met in view of the MRR for the at least some of the plurality of candidate/hypernym pairs.

20. The computer system of claim 19 , wherein in determining whether the predetermined loss threshold is met, the computing system is further configured to:

determine a Mean Average Precision (MAP) for at least some of the plurality of candidate/hypernym pairs; and

determine whether predetermined loss threshold is met in view of the MRR and the MAP for the at least some of the plurality of candidate/hypernym pairs.

Assignments (2)
SECURITY INTEREST Recorded May 29, 2020
From: PINTEREST, INC.
To: JPMORGAN CHASE BANK, N.A.
Reel/Frame 052797/0001 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 3, 2019
From: MANZOOR, EMAAD AHMED; LI, RUI; SHROUTY, DHANANJAY; LESKOVEC, JURIJ
To: PINTEREST, INC.
Reel/Frame 050618/0787 →
References Cited (99)
US 20190325342A1 · Sikka · 2019 [cited by examiner]
US 20200074246A1 · Goyal · 2020 [cited by examiner]
Shi, Y., Shen, J., Li, Y., Zhang, N., He, X., Lou, Z., . . . & Han, J. (2019). Discovering Hypernymy in Text-Rich Heterogeneous Information Network by Exploiting Context Granularity. arXiv preprint arXiv:1909.01584. (Ye… [cited by examiner]
Luu, A. T., Tay, Y., Hui, S. C., & Ng, S. K. (Nov. 2016). Learning term embeddings for taxonomic relation identification using dynamic weighting neural network. In Proceedings of the 2016 Conference on Empirical Methods… [cited by examiner]
You, J., Ying, R., & Leskovec, J. (2019). Position-aware Graph Neural Networks. arXiv e-prints, arXiv-1906. (Year: 2019). [cited by examiner]
Shen, Z., Ma, H., & Wang, K. (2018). A web-scale system for scientific knowledge exploration. arXiv preprint arXiv: 1805.12216. (Year: 2018). [cited by examiner]
Zhang, X., Chen, Y., Chen, J., Du, X., Wang, K., & Wen, J. R. (Aug. 2017). Entity set expansion via knowledge graphs. In Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Informat… [cited by examiner]
McFee, B., & Lanckriet, G. R. (2010, January). Metric learning to rank. In ICML. (Year: 2010). [cited by examiner]
McFee et al., “Metric Learning to Rank”, 2010, Proceedings of the 27th International Conference on Machine Learning, vol. 27 (2010 ), pp. 1-8 (Year: 2010). [cited by examiner]
Zhang et al., “TaxoGen: Unsupervised Topic Taxonomy Construction by Adaptive Term Embedding and Clustering”, 2018, In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, vol… [cited by examiner]
Qu et al., “Weakly-supervised Relation Extraction by Pattern-enhanced Embedding Learning”, 2018, Proceedings of the 2018 World Wide Web Conference, vol. 2018, pp. 1257-1266 (Year: 2018). [cited by examiner]
Qu et al., “Probabilistic Logic Neural Networks for Reasoning”, Jun. 20, 2019, arXiv, v1906.08495v1, pp. 1-10 (Year: 2019). [cited by examiner]
Shen et al., “HiExpan: Task-Guided Taxonomy Construction by Hierarchical Tree Expansion”, 2018, Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, vol. 24(2018), pp. 2180-2… [cited by examiner]
Shi et al., “ProjE: Embedding Projection for Knowledge Graph Completion”, 2017, Proceedings of the AAAI Conference on Artificial Intelligence, vol. 31(1), pp. 1236-1242 (Year: 2017). [cited by examiner]
Shi et al., “Discovering Hypernymy in Text-Rich Heterogeneous Information Network by Exploiting Context Granularity”, Sep. 4, 2019 , arXiv, v1909.01584v1, pp. 1-10 (Year: 2019). [cited by examiner]
Pennington, J., Socher, R. and Manning, C. 2014. Glove: Global Vectors for Word Representation. In EMNLP. https://hlp.stanford.edu/pubs/glove.pdf. [cited by applicant]
Perozzi, B., Al-Rfou, R. and Skiena, S. 2014. Deepwalk: Online Learning of Social Representations. In KDD. http://perozzi.net/publications/14_kdd_deepwalk.pdf. [cited by applicant]
Roller, S., Erk, K. and Boleda, G. 2014. Inclusive yet Selective: Supervised Distributional Hypemymy Detection. In COLING. http://www.cs.utexas.edu/users/ml/papers/roller.coling14.pdf. [cited by applicant]
Sala, F., De Sa, C., Gu, A. and Ré, C. 2018. Representation Tradeoffs for Hyperbolic Embeddings. In ICML. http://proceedings.mlr.press/v80/sala18a/sala18a.pdf. [cited by applicant]
Schroff, F., Kalenichenko, D. and Philbin, J. 2015. Facenet: A Unified Embedding for Face Recognition and Clustering. In CVPR. http://openaccess.thecvf.com/content_cvpr_2015/papers/Schroff_FaceNet_A_Unified_2015_CVPR_pa… [cited by applicant]
Shaw, B. and Jebara, T. 2009. Structure Preserving Embedding. In ICML. http://www.cs.columbia.edu/˜jebara/papers/spe-icm109.pdf. [cited by applicant]
Shaw, B., Huang, B., and Jebara, T. 2011. Learning a Distance Metric from A Network. In NIPS. http://www.cs.columbia.edu/˜jebara/papers/metricfromnetwork.pdf. [cited by applicant]
Shen, J., Wu, Z., Lei, D., Zhang, C., Ren, X., Vanni, M.T., Sadler, B.M. and Han, J. 2018. HiExpan: Task-Guided Taxonomy Construction by Hierarchical Tree Expansion. In SIGKDD. http://hanj.cs.illinois.edu/pdf/kdd18_jshe… [cited by applicant]
Shwartz, V., Goldberg, Y. and Dagan, I. 2016. Improving Hypernymy Detection with an Integrated Path-Based and Distributional Method. arXiv preprint arXiv:1603.06076 (2016). https://www.aclweb.org/anthology/P16-1226.pdf. [cited by applicant]
Snow, R., Jurafsky, D. and Ng, A.Y. 2005. Learning Syntactic Patterns for Automatic Hypernym Discovery. In NIPS. http:/ai.stanford.edu/˜rion/papers/hypernym_nips05.pdf. [cited by applicant]
Socher, R., Chen, D., Manning, C.D. and Ng, A. 2013. Reasoning with Neural Tensor Networks for Knowledge Base Completion. In NIPS. https://nlp.stanford.edu/pubs/SocherChenManningNg_NIPS2013.pdf. [cited by applicant]
Speer, R., Chin, J. and Havasi, C. 2017. ConceptNet 5.5: An Open Multilingual Graph of General Knowledge. In AAAI. https://tianjun.me/static/essay_resources/Introduction_to_ConceptNet/ConceptNet%205.5%20An%20Open%20Mult… [cited by applicant]
Tang, J., Qu, M., Wang, M., Zhang, M., Yan, J. and Mei, Q. 2015. Line: Large-Scale Information Network Embedding. In WWW. 1067-1077. http://www.www2015.it/documents/proceedings/proceedings/p1067.pdf. [cited by applicant]
Trouillon, T., Welbl, J., Riedel, S., Gaussier, E. and Bouchard, G. 2016. Complex Embeddings for Simple Link Prediction. In ICML. http://proceedings.mlr.press/v48/trouillon16.pdf. [cited by applicant]
Ustalov, D., Arefyev, N., Biemann, C. and Panchenko, A. 2017. Negative Sampling Improves Hypernymy Extraction Based on Projection Learning. arXiv preprint arXiv:1707.03903 (2017).https:/arxiv.org/pdf/1707.03903.pdf. [cited by applicant]
Vendrov, I., Kiros, R., Fidler, S. and Urtasun, R. 2015. Order-Embeddings of Images and Language. arXiv preprint arXiv:1511.06361 (2015). https://arxiv.org/pdf/1511.06361v3.pdf. [cited by applicant]
Verma, N., Mahajan, D., Sellamanickam, S. and Nair, V. 2012. Learning Hierarchical Similarity Metrics. In CVPR. http://www.cs.toronto.edu/˜vnair/cvpr12.pdf. [cited by applicant]
Vilnis, L. and McCallum, A. 2015. Word Representations via Gaussian Embedding. In ICLR. https://arxiv.org/pdf/1412.6623v3.pdf. [cited by applicant]
Vilnis, L., Li, X., Murty, S. and McCallum, A. 2018. Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures. arXiv preprint arXiv:1805.06627 (2018). https://arxiv.org/pdf/1805.06627.pdf. [cited by applicant]
Vulic, I., Glavaš, G., Mrkšić, N. and Korhonen, A. 2018. Post-Specialisation: Retrofitting Vectors of Words Unseen in Lexical Resources. arXiv preprint arXiv:1805.03228 (2018).https://arxiv.org/pdf/1805.03228.pdf. [cited by applicant]
Wang, C., He, X. and Zhou, A. 2017. A Short Survey on Taxonomy Learning from Text Corpora: Issues, Resources and Recent Advances. In EMNLP. https://chywang.github.io/papers/emnlp2017b.pdf. [cited by applicant]
Wang, J., Zhou, S., Wang, J. and Hou, Q. 2018. Deep Ranking Model by Large Adaptive Margin Learning for Person Re-Identification. Pattern Recognition (2018). [cited by applicant]
Weeds, J., Clarke, D., Reffin, J., Weir, D. and Keller, B. 2014. Learning to Distinguish Hypernyms and Co-Hyponyms. In COLING. https://www.researchgate.net/profile/Jeremy_Reffin/publication/273103581_Learning_to_Disting… [cited by applicant]
Weinberger, K.Q. and Chapelle, O. 2009. Large Margin Taxonomy Embedding for Document Categorization. In NIPS. https://www.researchgate.net/profile/Kilian_Weinberger/publication/228633511_Large_Margin_Taxonomy_Embedding_… [cited by applicant]
Weinberger, K.Q. and Saul, L.K. 2009. Distance Metric Learning for Large Margin Nearest Neighbor Classification. UMLR (2009). http://jmlr.csail.mit.edu/papers/volume10/weinberger09a/weinberger09a.pdf. [cited by applicant]
Wu, C-Y, Manmatha, R, Smola, A.J. and Krahenbuhl, P. 2017. Sampling Matters in Deep Embedding Learning. In CCV. https://arxiv.org/pdf/1706.07567v1.pdf. [cited by applicant]
Xing, E.P., Jordan, M.I., Russell, S.J. and Ng, A.Y. 2003. Distance Metric Learning with Application to Clustering with Side-Information. In NIPS. http://citeseer.ist.psu.edu/viewdoc/download;isessionid=AB5B7097335BD16D… [cited by applicant]
Yamane, J., Takatani, T., Yamada, H., Miwa, M. and Sasaki, Y. 2016. Distributional Hypernym Generation by Jointly Learning Clusters and Projections. In COLING. https://www.aclweb.org/anthology/C16-1176.pdf. [cited by applicant]
Yang, H. and Callan, J. 2009. A Metric-Based Framework for Automatic Taxonomy Induction. In ACL. http://www.cs.cmu.edu/˜callan/Papers/acl09-huiyang.pdf. [cited by applicant]
Yu, Z., Wang, H., Lin, X. and Wang, M. 2015. Learning Term Embeddings for Hypernymy Identification. In IJCAI. https://www.ijcai.org/Proceedings/15/Papers/200.pdf. [cited by applicant]
Zhang, C., Tao, F., Chen, X., Shen, J., Jiang, M., Sadler, B., Vanni, M. and Han, J. 2018. Taxogen: Unsupervised Topic Taxonomy Construction by Adaptive Term Embedding and Clustering. In SIGKDD. https://research.fb.com/… [cited by applicant]
Zhang, Y., Ahmed, A., Josifovski, V. and Smola, A. 2014. Taxonomy Discovery for Personalized Recommendation. In WSDM. https://zhangyuc.github.io/files/zhang14wsdm.pdf. [cited by applicant]
Zhuang, J. and Liu, Y. 2019. PinText: A Multitask Text Embedding System in Pinterest. In KDD. https://dl.acm.org/doi/pdf/10.1145/3292500.3330671?download=true. [cited by applicant]
Ziegler, C-N, Lausen, G., and Schmidt-Thieme, L. 2004. Taxonomy-Driven Computation of Product Recommendations. In CIKM. https://www.researchgate.net/profile/Cai-Nicolas_Ziegler/ publication/200121010_Taxonomy-driven_Com… [cited by applicant]
Agrawal, R., Gollapudi, S., Halverson, A. and leong, S. 2009. Diversifying Search Results. In WSDM. [cited by applicant]
Ahmed, N.K., Rossi, R., Lee, J.B., Willke, T.L., Zhou, R., Kong, X. and Eldardiry, H. 2018. Learning Role-based Graph Embeddings. arXiv preprint arXiv:1802.02896 (2018). [cited by applicant]
Aly, R., Acharya, S., Ossa, A., Kohn, A., Biemann, C. and Panchenko, A. 2019. Every Child Should Have Parents: A Taxonomy Refinement Algorithm Based on Hyperbolic Term Embeddings. In ACL. [cited by applicant]
Anh Tuan, .L., Tay, Y., Hui, S.C. and Ng, S.K. 2016. Learning Term Embeddings for Taxonomic Relation Identification Using Dynamic Weighting Neural Network. In EMNLP. 403-413. [cited by applicant]
Babbar, R., Partalas, I., Gaussier, E. and Amini, M.R. 2013. On Flat Versus Hierarchical Classification in Large-scale Taxonomies. In NIPS. http://papers.nips.cc/paper/5082-on-flat-versus-hierarchical-classification-in-… [cited by applicant]
Babbar, R., Partalas, I., Gaussier, E. and Amini, M-R, and Amblard, C. 2016. Learning Taxonomy Adaptation in Large- scale Classification. JMLR (2016). http://jmlr.org/papers/volume17/14-207/14-207.pdf. [cited by applicant]
Bansal, M., Burkett, D., De Melo, G. and Klein, D. 2014. Structured Learning for Taxonomy Induction with Belief Propagation. In ACL. https://www.researchgate.net/publication/270877287_Structured_Learning_for_Taxonomy_In… [cited by applicant]
Baroni, M., Bernardi, R., Do, N-Q, and Shan, C-C. 2012. Entailment Above the Word Level in Distributional Semantics. In EACL. https://www.researchgate.net/publication/228438454_Entailment_above_the_word_level_in_distrib… [cited by applicant]
Bernier-Colborne, G. and Barriere, C. 2018. CRIM at SemEval-2018 Task 9: A Hybrid Approach to Hypernym Discovery. In Proceedings of the 12th International Workshop on Semantic Evaluation. https://www.aclweb.org/antholog… [cited by applicant]
Bojanowski, P., Grave, E., Joulin, A. and Mikolov, T. 2017. Enriching Word Vectors with Subword Information. TACL (2017). https://www.mitpressjournals.org/doi/pdfplus/10.1162/tacl_a_00051. [cited by applicant]
Bordes, A., Usunier, N., Garcia-Duran, A., Weston, J., and Yakhnenko, O. 2013. Translating Embeddings for Modeling Multi-relational Data. In NIPS. http://papers.nips.cc/paper/5071-translating-embeddings-for-modeling-mul… [cited by applicant]
Camacho-Collados, J., Delli Bovi, C., Anke, L.E., Oramas, S., Pasini, T., Santus, E., Shwartz, V., Navigli, R., and Saggion, H. 2018. SemEval-2018 task 9: Hypernym Discovery. In Proceedings of the 12th International Wor… [cited by applicant]
Dekel, O., Keshet, J., and Singer, Y. 2004. Large Margin Hierarchical Classification. In ICML.https://oferdekel.github.io/pdf/2004DekelKeSi.pdf. [cited by applicant]
Dong, T., Bauckhage, C., Jin, H., Li, J., Cremers, O., Speicher, D., Cremers, A.B., and Zimmermann, J. 2018. Imposing Category Trees onto Word-Embeddings Using a Geometric Construction. (2018). https://www.researchgate.… [cited by applicant]
Donnat, C., Zitnik, M., Hallac, D., and Leskovec, J. 2018. Learning Structural Node Embeddings via Diffusion Wavelets. In KDD. https://cs.stanford.edu/˜jure/pubs/graphwave-kdd18.pdf. [cited by applicant]
Eksombatchai, C., Jindal, P., Liu, J.Z., Liu, Y., Sharma, R., Sugnet, C., Ulrich, M. and Leskovec, J. 2018. Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time. In WWW. https://cs.stanfo… [cited by applicant]
Espinosa-Anke, L., Camacho-Collados, J., Delli Bovi, C., and Saggion, H. 2016. Supervised Distributional Hypernym Discovery via Domain Adaptation. In EMNLP. https://www.researchgate.net/publication/309726823_Supervised_… [cited by applicant]
Feng, Y., Wang, H., Yi, D.T. and Hu, R.2019. Triplet Distillation for Deep Face Recognition. https://arxiv.org/pdf/1905.04457.pdf (2019). [cited by applicant]
Figueiredo, D.R., Ribeiro, L.FR, and Saverese, P.HP, 2017. struc2vec: Learning Node Representations from Structural Identity. In KDD. https://arxiv.org/pdf/1704.03165v1.pdf. [cited by applicant]
Fu, R., Guo, J., Qin, B., Che, W., Wang, H. and Liu, T. 2014. Learning Semantic Hierarchies via Word Embeddings. In ACL. http://ir.hit.edu.cn/˜jguo/papers/acl2014-hypernym.pdf. [cited by applicant]
Ganea, O-E, Bécigneul, G. and Hofmann, T. 2018. Hyperbolic Entailment Cones for Learning Hierarchical Embeddings. https://arxiv.org/pdf/1804.01882.pdf (2018). [cited by applicant]
Geffet, M. and Dagan, I. 2005. The Distributional Inclusion Hypotheses and Lexical Entailment. In ACL. http://ssli.ee.washington.edu/conferences/ACL2005/ACL/pdf/ACL14.pdf. [cited by applicant]
Gonçalves, R.S., Horridge, M., Li, R., Liu, Y., Musen, M.A., Nyulas, C.I., Obamos, E., Shrouty, D., and Temple, D. 2019. Use of OWL and Semantic Web Technologies at Pinterest. https://arxiv.org/pdf/1907.02106.pdf(2019). [cited by applicant]
Grover, A. and Leskovec, J. 2016. node2vec: Scalable Feature Learning for Networks. In KDD. [cited by applicant]
He, R., Lin, C., Wang, J. and McAuley, J. 2016. Sherlock: Sparse Hierarchical Embeddings for Visually-Aware One-Class Collaborative Filtering. In AAAl. https://cseweb.ucsd.edu/˜jmcauley/pdfs/ijcai16.pdf. [cited by applicant]
Hearst, M.A. 1992. Automatic Acquisition of Hyponyms from Large Text Corpora. In ACL. https://www.aclweb.org/anthology/C92-2082.pdf. [cited by applicant]
Huang, J., Ren, Z., Zhao, W.X., He, G., Wen, J-R, and Dong, D. 2019. Taxonomy-Aware Multi-Hop Reasoning Networks for Sequential Recommendation. In WSDM. [cited by applicant]
Kanagal, B., Ahmed, A., Pandey, S., Josifovski, V., Yuan, J. and Garcia-Pueyo, L. 2012. Supercharging Recommender Systems Using Taxonomies for Learning User Purchase Behavior. VLDB (2012). http://vldb.org/pvldb/vol5/p95… [cited by applicant]
Kingma, D.P. and Ba, J. 2014. Adam: A Method for Stochastic Optimization. arXiv preprint arXiv:1412.6980 (2014). https://arxiv.org/pdf/1412.6980v1.pdf. [cited by applicant]
Laha, A., Chemmengath, S.A., Agrawal, P., Khapra, M., Sankaranarayanan, K. and Ramaswamy, H.G. 2018. On Controllable Sparse Alternatives to Softmax. In NIPS. https://arxiv.org/pdf/1810.11975.pdf. [cited by applicant]
Levy, O., Remus, S., Biemann, C. and Dagan, I. 2015. Do Supervised Distributional Methods Really Learn Lexical Inference Relations? In NAACL-HLT. https://www.aclweb.org/anthology/N15-1098.pdf. [cited by applicant]
Li, X., Vilnis, L., Zhang, D., Boratko, M. and McCallum, A. 2018. Smoothing the Geometry of Probabilistic Box Embeddings. (2018). https://openreview.net/pdf?id=H1xSNiRcF7. [cited by applicant]
Lim, D. and Lanckriet, G. 2014. Efficient Learning of Mahalanobis Metrics for Ranking. In ICML.http://proceedings.mlr.press/v32/lim14.pdf. [cited by applicant]
Liu, D., Rogers, S., Shiau, R., Kislyuk, D., Ma, K.C., Zhong, Z., Liu, J. and Jing, Y. 2017. Related Pins at Pinterest: The Evolution of a Real-World Recommender System. In WWW. https://arxiv.org/pdf/1702.07969v1.pdf. [cited by applicant]
Liu, N., Huang, X., Li, J. and Hu, X. 2018. On Interpretation of Network Embedding Via Taxonomy Induction. In KDD. http://www.public.asu.edu/˜jundongl/paper/KDD18_Network_Embedding_Interpretation.pdf. [cited by applicant]
Mao, Y., Ren, X., Shen, J., Gu, X. and Han, J. 2018. End-to-End Reinforcement Learning for Automatic Taxonomy Induction. In SDM. https://arxiv.org/pdf/1805.04044.pdf. [cited by applicant]
Martins, A. and Astudillo, R. 2016. From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification. In ICML. https://arxiv.org/pdf/1602.02068v2.pdf. [cited by applicant]
McCallum, A., Rosenfeld, R., Mitchell, T.M. and Ng, A.Y. 1998. Improving Text Classification by Shrinkage in a Hierarchy of Classes. In ICML. https://pdfs.semanticscholar.org/f267/1b151fad7e176176b35d425b2b6356ff4595.pd… [cited by applicant]
McFee, B. and Lanckriet, G.R. 2010. Metric Learning to Rank. In ICML. https://bmcfee.github.io/papers/mlr.pdf. [cited by applicant]
Menon, A.K., Chitrapura, K-P, Garg, S., Agarwal, D. and Kota, N. 2011. Response Prediction Using Collaborative Filtering with Hierarchies and Side-Information. In KDD. http://ideal.ece.utexas.edu/courses/ee380l_addm/kdd… [cited by applicant]
Mikolov, T., Sutskever, I., Chen, K., Corrado, G.S. and Dean, J. 2013. Distributed Representations of Words and Phrases and Their Compositionality. In NIPS. https://arxiv.org/pdf/1310.4546.pdf. [cited by applicant]
Mikolov, T., Yih, W-T and Zweig, G. 2013. Linguistic Regularities in Continuous Space Word Representations. In NAACL-HLT. https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/rvecs.pdf. [cited by applicant]
Miller, G.A. 1995. WordNet: A Lexical Database for English. CACM (1995). http://12r.cs.uiuc.edu/Teaching/CS598-05/Papers/miller95.pdf. [cited by applicant]
Mirzazadeh, F., Guo, Y. and Schuurmans, D. 2014. Convex Co-Embedding. In AAAl. https://researcher.watson.ibm.com/researcher/files/IBM-Farzaneh/14_aaai14paper.pdf. [cited by applicant]
Navigli, R. and Velardi, P. 2010. Learning Word-Class Lattices for Definition and Hypernym Extraction. In ACL.https://www.aclweb.org/anthology/P10-1134.pdf. [cited by applicant]
Nguyen, K.A., Koper, M. Schulte im Walde, S., and Vu, N.T. 2017. Hierarchical Embeddings For Hypernymy Detection and Directionality. arXiv preprint arXiv:1707.07273 (2017).https://arxiv.org/pdf/1707.07273.pdf. [cited by applicant]
Nickel, M. and Kiela, D. 2017. Poincare Embeddings for Learning Hierarchical Representations. In NIPS. https://arxiv.org/pdf/1705.08039.pdf. [cited by applicant]
Nickel, M. and Kiela, D. 2018. Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometry. In ICML. https://arxiv.org/pdf/1806.03417.pdf. [cited by applicant]
Nickel, M., Rosasco, L. and Poggio, T. 2016. Holographic Embeddings of Knowledge Graphs. In AAAl. https://cbmm.mit.edu/sites/default/files/publications/1510.04935v2.pdf. [cited by applicant]
Paccanaro, A. and Hinton, G.E. 2002. Learning Hierarchical Structures with Linear Relational Embedding. In NIPS. http://papers.nips.cc/paper/2068-learning-hierarchical-structures-with-linear-relational-embedding.pdf. [cited by applicant]
Cited By (1)
US 12,608,733