Improving word embeddings projection for Turkish hypernym extraction

Basit Öğe Kaydı
dc.WoS.categoriesComputer Science, Artificial Intelligence; Engineering, Electrical & Electronicen_US
dc.authorid0000-0002-7764-2891en_US
dc.contributor.authorYıldırım, Savaş
dc.date.accessioned2021-02-18T06:52:36Z
dc.date.available2021-02-18T06:52:36Z
dc.date.issued2019-01
dc.description.abstractCorpus-driven approaches can automatically explore is-a relations between the word pairs from corpus. This problem is also called hypernym extraction. Formerly, lexico-syntactic patterns have been used to solve hypernym relations. The language-specific syntactic rules have been manually crafted to build the patterns. On the other hand, recent studies have applied distributional approaches to word semantics. They extracted the semantic relations relying on the idea that similar words share similar contexts. Former distributional approaches have applied one-hot bag-of-word (BOW) encoding. The dimensionality problem of BOW has been solved by various neural network approaches, which represent words in very short and dense vectors, or word embeddings. In this study, we used word embeddings representation and employed the optimized projection algorithm to solve the hypernym problem. The supervised architecture learns a mapping function so that the embeddings (or vectors) of word pairs that are in hypernym relations can be projected to each other. In the training phase, the architecture first learns the embeddings of words and the projection function from a list of word pairs. In the test phase, the projection function maps the embeddings of a given word to a point that is the closest to its hypernym. We utilized the deep learning optimization methods to optimize the model and improve the performances by tuning hyperparameters. We discussed our results by carrying out many experiments based on cross-validation. We also addressed problem-specific loss function, monitored hyperparameters, and evaluated the results with respect to different settings. Finally, we successfully showed that our approach outperformed baseline functions and other studies in the Turkish language.en_US
dc.fullTextLevelFull Texten_US
dc.identifier.doi10.3906/elk-1903-65
dc.identifier.issn1303-6203
dc.identifier.issn1300-0632
dc.identifier.scopus2-s2.0-85076641356en_US
dc.identifier.trdizinid337966en_US
dc.identifier.urihttps://hdl.handle.net/11411/3259
dc.identifier.urihttps://doi.org/10.3906/elk-1903-65
dc.identifier.urihttps://search.trdizin.gov.tr/yayin/detay/337966en_US
dc.identifier.wosWOS:000506165400028en_US
dc.identifier.wosqualityQ4en_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.indekslendigikaynakTR-Dizinen_US
dc.issue6en_US
dc.language.isoenen_US
dc.nationalInternationalen_US
dc.numberofauthors1en_US
dc.pages4418-4428en_US
dc.publisherTUBITAK SCIENTIFIC & TECHNICAL RESEARCH COUNCIL TURKEYen_US
dc.relation.ispartofTURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCESen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectProjection learningen_US
dc.subjectword embeddingsen_US
dc.subjecthypernym relationen_US
dc.subjectdeep learningen_US
dc.titleImproving word embeddings projection for Turkish hypernym extraction
dc.typeArticle
dc.volume27en_US

Dosyalar

Orijinal paket
Listeleniyor 1 - 1 / 1
Küçük Resim
İsim:
2019WosYıldırım.pdf
Boyut:
1.99 MB
Biçim:
Adobe Portable Document Format
Açıklama:
İndir
Lisans paketi
Listeleniyor 1 - 1 / 1
Küçük Resim Yok
İsim:
license.txt
Boyut:
1.71 KB
Biçim:
Item-specific license agreed upon to submission
Açıklama:
İndir

Koleksiyon

Faculty of Engineering and Natural Sciences
Scopus Indexed Publications
TR Dizin Indexed Publications
Web of Science Indexed Publications