?
Evaluation of Vector Transformations for Russian Word2Vec and FastText Embeddings
.
Authors of Word2Vec claimed that their technology could solve the word analogy problem using the vector transformation in the introduced vector space. However, the practice demonstrates that it is not always true. In this paper, we investigate several Word2Vec and FastText model trained for the Russian language and find out reasons of such inconsistency. We found out that different types of words are demonstrating different behavior in the semantic space. FastText vectors are tending to find phonological analogies, while Word2Vec vectors are better in finding relations in geographical proper names. However, we found out that just four out of fifteen selected domains are demonstrating accuracy more that 0.8. We also draw a conclusion that in a common case, the task of word analogies could not be solved using a random word pair taken from two investigated categories. Our experiments have demonstrated that in some cases the length of the vectors could differ more than twice. Calculation of an average vector leads to a better solution here since it closer to more vectors.
In book
St. Petersburg: CEUR-WS, 2020.
Glavas G., Franco-Salvador M., Ponzetto S. et al., Knowledge-Based Systems 2018 Vol. 143 P. 1–9
Recognizing semantically similar sentences or paragraphs across languages is beneficial for many tasks, ranging from cross-lingual information retrieval and plagiarism detection to machine translation. Recently proposed methods for predicting cross-lingual semantic similarity of short texts, however, make use of tools and resources (e.g., machine translation systems, syntactic parsers or named entity recognition) that for many ...
Added: October 29, 2020
Gharavi E., Veisi H., Россо П., Neural Computing and Applications 2020 Vol. 32 No. 14 P. 10593–10607
The efficiency and scalability of plagiarism detection systems have become a major challenge due to the vast amount of available textual data in several languages over the Internet. Plagiarism occurs in different levels of obfuscation, ranging from the exact copy of original materials to text summarization. Consequently, designed algorithms to detect plagiarism should be robust ...
Added: October 29, 2020
Arefyev N V., Fedoseev M., Kabanov A. et al., , in: Компьютерная лингвистика и интеллектуальные технологии: по материалам ежегодной международной конференции «Диалог» (Москва, 17–20 июня 2020 г.)Issue 19(26): дополнительный том.: -, 2020. P. 13–32.
Expert-built lexical resources are known to provide information of good quality for the cost of low coverage. This property limits their applicability in modern NLP applications. Building descriptions of lexical-semantic relations manually in sufficient volume requires a huge amount of qualified human labour. However, given some initial version of a taxonomy is already built, automatic ...
Added: October 9, 2020
Arefyev, N., Ermolaev P., Panchenko A., , in: Computational Linguistics and Intellectual Technologies. International Conference "Dialogue 2018" Proceedings.: M.: Conference Proceedings Editorial board, 2018. P. 68–84.
The paper describes our participation in the first shared task on word sense induction and disambiguation for the Russian language RUSSE'2018 [Panchenko et al., 2018]. For each of several dozens of ambiguous words, the participants were asked to group text fragments containing it according to the senses of this word, which were not provided beforehand, ...
Added: October 9, 2020
Karyaeva M., Braslavski P., Sokolov V., Automatic Control and Computer Sciences 2019 Vol. 53 P. 638–643
The ability to identify semantic relations between words has made a word2vec model widely used in NLP tasks. The idea of word2vec is based on a simple rule that a higher similarity can be reached if two words have a similar context. Each word can be represented as a vector, so the closest coordinates of vectors can be interpreted ...
Added: April 10, 2020
Ekaterina Artemova, Bakarov A., Artemov A. et al., Journal of Cognitive Science 2020 Vol. 1 No. 21 P. 15–52
In this paper, our focus is the connection and influence of language technologies on the research in neurolinguistics. We present a review of brain imaging-based neurolinguistics studies with a focus on the natural language representations, such as word embeddings and pre-trained language model. Mutual enrichment of neurolinguistics and language technologies leads to development of brain-aware natural ...
Added: January 17, 2020
Zobnin A., Elistratova E., , in: Proceedings of the 4th Workshop on Representation Learning for NLP (RepL4NLP-2019)Issue W19-43.: Association for Computational Linguistics, 2019. P. 244–249.
Most word embedding algorithms such as word2vec or fastText construct two sort of vectors: for words and for contexts. Naive use of vectors of only one sort leads to poor results. We suggest using indefinite inner product in skip-gram negative sampling algorithm. This allows us to use only one sort of vectors without loss of ...
Added: November 9, 2019
Puzyrev D., Shelmanov A., Panchenko A. et al., , in: Proceedings of the 7th Workshop on Balto-Slavic Natural Language Processing, 2019, Florence, Italy, Association for Computational Linguistics.: Association for Computational Linguistics, 2019. P. 56–62.
aper presents the first gold-standard resource for Russian annotated with compositionality information of noun compounds. The compound phrases are collected from the Universal Dependency treebanks according to part of speech patterns, such as ADJ+NOUN or NOUN+NOUN, using the gold-standard annotations. Each compound phrase is annotated by two experts and a moderator according to the following ...
Added: October 30, 2019
Puzyrev D. A., Shelmanov A., Panchenko A. et al., , in: Analysis of Images, Social Networks and Texts. 8th International Conference AIST 2019.: Springer, 2019. P. 218–229.
In this paper, we present the first gold-standard corpus of Russian noun compounds annotated with compositionality information. We used Universal Dependency treebanks to collect noun compounds according to part of speech patterns, such as ADJ-NOUN or NOUN-NOUN and annotated them according to the following schema: a phrase can be either compositional, non-compositional, or ambiguous (i.e., ...
Added: October 30, 2019
Toldova S., Pisarevskaya D., Kobozeva M., , in: Artificial Intelligence and Natural Language, 7th International Conference, AINL 2018, St. Petersburg, Russia, October 17–19, 2018, ProceedingsIssue 930.: Switzerland: Springer, 2018. P. 79–87.
The identification of discourse connectives plays an important role in many discourse processing approaches. Among them there are functional words usually enumerated in grammars (iz-za ‘due to’, blagodarya ‘thanks to’,) and not grammaticalized expressions (X vedet k Y ‘X leads to Y’, prichina etogo ‘the cause is’). Both types of connectives signal certain relations between ...
Added: October 26, 2018
Voronkov Ilia, Amajd M., Kaimuldenov Z., , in: Actual Problems of System and Software Engineering 2017. Proceedings of the 5th International Conference on Actual Problems of System and Software Engineering Supported by Russian Foundation for Basic Research. Project #17-07-20565 Moscow, Russia, November 14-16, 2017, 408 P.Vol. 1989.: Aachen: CEUR Workshop Proceedings, 2017. P. 362–370.
In this paper, we analyze the use of different neural networks for the
text classification task. The accuracy of the studied text classifiers can be
changed by a small number of previously classified texts. This is important due
to the fact that in many applications of text classification a large number of unlabeled texts are easily accessible, while ...
Added: August 16, 2018
Zobnin A., , in: Analysis of Images, Social Networks and Texts. 6th International Conference, 2017, Revised Selected PapersVol. 10716.: Cham: Springer, 2018. Ch. 11 P. 116–128.
Consider a continuous word embedding model. Usually, the cosines between word vectors are used as a measure of similarity of words. These cosines do not change under orthogonal transformations of the embedding space. We demonstrate that, using some canonical orthogonal transformations from SVD, it is possible both to increase the meaning of some components and ...
Added: November 26, 2017
Wohlgenannt G., Artemova E., Ilvovsky D., , in: Proceedings of the Workshop on Language Technology Resources and Tools for Digital Humanities (LT4DH).: Osaka: [б.и.], 2016. Ch. 4 P. 18–26.
In this paper a social network is extracted from a literary text. The social network shows, how frequent the characters interact and how similar their social behavior is. Two types of similarity measures are used: the first applies co-occurrence statistics, while the second exploits cosine similarity on different types of word embedding vectors. The results ...
Added: March 6, 2017
Kreps V. L., International Game Theory Review 2017 Vol. 19 No. 2
We consider finite noncooperative (Formula presented.) person games with fixed numbers (Formula presented.), (Formula presented.), of pure strategies of Player (Formula presented.). We propose the following question: is it possible to extend the vector space of finite noncooperative (Formula presented.)-games in mixed strategies such that all games of a broader vector space of noncooperative (Formula ...
Added: February 20, 2017
Kutuzov A. B., Kuzmenko E., Marakasova A., , in: Proceedings of the Workshop on Language Technology Resources and Tools for Digital Humanities (LT4DH).: Osaka: [б.и.], 2016. P. 26–34.
We present an approach to detect differences in lexical semantics across English language registers, using word embedding models from distributional semantics paradigm. Models trained on register-specific subcorpora of the BNC corpus are employed to compare lists of nearest associates for particular words and draw conclusions about their semantic shifts depending on register in which they ...
Added: November 12, 2016
Kutuzov A. B., Velldal E., Øvrelid L., , in: Proceedings of The 20th SIGNLL Conference on Computational Natural Language Learning.: Berlin: Association for Computational Linguistics, 2016. P. 115–125.
This paper studies how word embeddings trained on the British National Corpus interact with part of speech boundaries. Our work targets the Universal PoS tag set, which is currently actively being used for annotation of a range of languages. We experiment with training classifiers for predicting PoS tags for words based on their embeddings. The ...
Added: November 12, 2016
Kutuzov A. B., Козлова О. С., , in: Компьютерная лингвистика и интеллектуальные технологии: По материалам ежегодной международной конференции «Диалог» (Москва,1–4 июля 2016 г.)Вып. 15.: М.: Изд-во РГГУ, 2016. P. 288–300.
In natural language processing, distributional semantic models are known as an efficient data driven approach to word and text representation, which allows computing meaning directly from large text corpora into word embeddings in a vector space. This paper addresses the role of linguistic preprocessing in enhancing performance of distributional models, and particularly studies pronominal anaphora ...
Added: November 12, 2016
Kreps V. L., / NRU Higher School of Economics. Series WP BRP "Economics/EC". 2016. No. WP BRP 150/EC/2016.
We consider finite non-cooperative N person games with fixed numbers mi, i = 1, . . . , N , of pure strategies of player i. We propose the following question: is it possible to extend the vector space of finite non-cooperative m1 × m2 × . . . × mN - games in mixed ...
Added: September 22, 2016