Search

Your search keyword '"Andrey Kutuzov"' showing total 91 results
Start Over Author "Andrey Kutuzov"
91 results on '"Andrey Kutuzov"'

47. Contextualized embeddings for semantic change detection: Lessons learned

Author

Andrey Kutuzov, Erik Velldal, and Lilja Øvrelid

Subjects
General Medicine
Abstract

This paper presents a qualitative analysis of the outputs of contextualized embedding-based methods for detecting diachronic semantic change. We also introduce an ensemble-method outperforming previously described contextualized approaches. This method is used as a basis for an in-depth analysis of the degrees of semantic change predicted for English words across 5 decades. Our findings show that contextualized methods can sometimes predict high change scores for words which are not undergoing any real diachronic semantic shift. Such cases are discussed in detail with examples, and their linguistic categorization is proposed. Contextualized embeddings are prone to confound changes in lexicographic senses and changes in contextual variance, which naturally stems from their distributional nature. Additionally, they often merge together syntactic and semantic aspects of lexical entities. Finally, we propose possible solutions to these issues.

Published
2022

48. Unreasonable Effectiveness of Rule-Based Heuristics in Solving Russian SuperGLUE Tasks

Author

Tatyana Iazykova, Olga Bystrova, Andrey Kutuzov, and Denis Kapelyushnik

Subjects
FOS: Computer and information sciences, Computer Science - Computation and Language, Exploit, business.industry, Computer science, Natural language understanding, Rule-based system, computer.software_genre, Machine learning, Modern language, Benchmark (computing), Artificial intelligence, Language model, Heuristics, business, Set (psychology), Computation and Language (cs.CL), computer
Abstract

Leader-boards like SuperGLUE are seen as important incentives for active development of NLP, since they provide standard benchmarks for fair comparison of modern language models. They have driven the world's best engineering teams as well as their resources to collaborate and solve a set of tasks for general language understanding. Their performance scores are often claimed to be close to or even higher than the human performance. These results encouraged more thorough analysis of whether the benchmark datasets featured any statistical cues that machine learning based language models can exploit. For English datasets, it was shown that they often contain annotation artifacts. This allows solving certain tasks with very simple rules and achieving competitive rankings. In this paper, a similar analysis was done for the Russian SuperGLUE (RSG), a recently published benchmark set and leader-board for Russian natural language understanding. We show that its test datasets are vulnerable to shallow heuristics. Often approaches based on simple rules outperform or come close to the results of the notorious pre-trained language models like GPT-3 or BERT. It is likely (as the simplest explanation) that a significant part of the SOTA models performance in the RSG leader-board is due to exploiting these shallow heuristics and that has nothing in common with real language understanding. We provide a set of recommendations on how to improve these datasets, making the RSG leader-board even more representative of the real progress in Russian NLU., Comment: Accepted to Dialogue'2021

Published
2021

49. Analysis of Images, Social Networks and Texts : 11th International Conference, AIST 2023, Yerevan, Armenia, September 28–30, 2023, Revised Selected Papers

Author

Dmitry I. Ignatov, Michael Khachay, Andrey Kutuzov, Habet Madoyan, Ilya Makarov, Irina Nikishina, Alexander Panchenko, Maxim Panov, Panos M. Pardalos, Andrey V. Savchenko, Evgenii Tsymbalov, Elena Tutubalina, Sergey Zagoruyko, Dmitry I. Ignatov, Michael Khachay, Andrey Kutuzov, Habet Madoyan, Ilya Makarov, Irina Nikishina, Alexander Panchenko, Maxim Panov, Panos M. Pardalos, Andrey V. Savchenko, Evgenii Tsymbalov, Elena Tutubalina, and Sergey Zagoruyko

Subjects
Data mining, Machine learning, Database management, Natural language processing (Computer science), Information storage and retrieval systems, Application software
Abstract

This book constitutes revised selected papers from the thoroughly refereed proceedings of the 11th International Conference on Analysis of Images, Social Networks and Texts, AIST 2023, held in Yerevan, Armenia, during September 28-30, 2023. The 24 full papers included in this book were carefully reviewed and selected from 93 submissions. They were organized in topical sections as follows: natural language processing; computer vision; data analysis and machine learning; network analysis; and theoretical machine learning and optimization. The book also contains one invited talk in full paper length.

Published
2024

50. Recent Trends in Analysis of Images, Social Networks and Texts : 11th International Conference, AIST 2023, Yerevan, Armenia, September 28–30, Revised Selected Papers

Author

Dmitry I. Ignatov, Michael Khachay, Andrey Kutuzov, Habet Madoyan, Ilya Makarov, Irina Nikishina, Alexander Panchenko, Maxim Panov, Panos M. Pardalos, Andrey V. Savchenko, Evgenii Tsymbalov, Elena Tutubalina, Sergey Zagoruyko, Dmitry I. Ignatov, Michael Khachay, Andrey Kutuzov, Habet Madoyan, Ilya Makarov, Irina Nikishina, Alexander Panchenko, Maxim Panov, Panos M. Pardalos, Andrey V. Savchenko, Evgenii Tsymbalov, Elena Tutubalina, and Sergey Zagoruyko

Subjects
Data mining, Artificial intelligence, Application software, Database management, Social sciences—Data processing, Image processing—Digital techniques, Computer vision
Abstract

​This book constitutes the refereed proceedings of the 11th International Conference on Recent Trends in Analysis of Images, Social Networks and Texts, AIST 2023, held in Yerevan, Armenia, during September 28–30, 2023. The 19 full papers 2 short papers and 1 demo paper included in this book were carefully reviewed and selected from 52 submissions. They were organized in topical sections as follows: Natural Language Processing; Computer Vision; Data Analysis and Machine Learning; Network Analysis; Theoretical Machine Learning and Optimization; and Demo Paper.

Published
2024

Catalog

Books, media, physical & digital resources
See catalog results