Sentiment analysis is one of the most demanded natural language processing operations for solving applied problems. One of the key methods of automated sentiment analysis is supervised machine learning. In the presence of a large selection of ready-made solutions for determining the tonality, the results of the models give significant errors due to the complexity and contextual conditionality of the linguistic explication of emotions. The article presents the results of the validation of 6 models for determining the sentiment of Russian-language publications using a research validation dataset – expertly marked 300 statements extracted from social network messages on the subject of quality of life and corresponding to one of the sentiment types: positive, negative, neutral. To evaluate the performance of the models, interannotator agreement coefficients were used, in particular, Krippendorff’s alpha, Cohen’s kappa and Fleiss’s kappa coefficients. The obtained values of the coefficients showed a low level of reliability between the expert labels and the labels that were assigned by the models. Among the experiments performed, the lowest agreement coefficients were achieved for the Blanchefort model trained on Rusentiment data, and the highest for the model of the same developer trained on medical feedback data. Based on the results obtained, conclusions were drawn about the most common causes of disagreements in determining sentiment by machine learning models. Machine learning models correctly identify the tone of texts if they contain bright lexical markers that match in tone the general tone of the statement. On the contrary, problems in determining the tone of an emotionally charged message by the model are provoked by the presence of a word with the opposite tone in it. The use of emotive vocabulary that does not match the tone of the entire statement, the presence of marker words not in their direct meanings, the use of uppercase, forms of complicated communication (including irony, sarcasm) remain risk factors for attracting automated analysis resources: with a high degree of probability the automatic classification model will not be able to correctly determine the tone of the text. The main reason for the “difficulties” of the automated determination of sentiment is the complexity of the task of focusing on the utterance as an integral unit and the refusal to focus on individual formal indicators. The utterance is the minimum communicative unit of speech. Capturing its semantic and emotionally expressive integrity is a super task for machine learning models in sentiment analysis. So, it is still quite difficult to trust machine learning models in solving such a complex task as automated categorization of emotions. It is advisable to associate the prospects for research directions in this area, first of all, with the development of high-quality, linguistically sound training datasets.
The article presents the results of a study of narratives about national heroes in the historical communities of the most popular Russian social network VK. The study is based on 332,781 unique text messages extracted by automated methods using the VK open API. For data processing, the nodes of the PolyAnalyst analytical platform are used: the search query, sentiment analysis, text clustering, summaries, visualization nodes, and a number of nodes for text preprocessing. The data was interpreted on the base of the theoretical and methodological provisions of modern narratoLogy, according to which any communication process can be represented as a form of narration. The authors argue that historical personalities are regularly included in modern national contexts as a component of a single sign space that determines national identity and self-consciousness. Among the most popular are Vladimir Lenin (17,092 references), Josef Stalin (13,830 references), Nicholas II (4,917 references). The main narratives associated with Lenin relate to his faith in the struggle to improve people's Life and the foundation of Soviet Union, later marked by the Victory in the Great Patriotic War. Lenin's name remains significant at the everyday Level. Opinions about Stalin are traditionally controversial. The Largest cluster of texts about StaLin reveals the narrative of the Great Patriotic War and the man who Led his country to the Great Victory. The family narrative dominates in the texts about Nicholas II. For some modern Runet users, the Last Russian emperor is a symbol of “Russia that we have lost" Statistically significant are also narratives about NichoLas II as a saint, a martyr, a bad ruLer, and about renunciation. The appLied analysis of unstructured big data of sociaL media aLLows soLving fundamentaLLy new tasks of identifying new features of the mass historicaL consciousness, which expands our understanding of the historical memory space and specific characteristics of the nationaL identity of modern Russia.
Томский государственный педагогический университет, ТомскВведение. При изучении полисемантической лексики в синтагматическом аспекте отдельно необходимо говорить о специфике ее функционирования в художественной речи. Если в контексте обыденной речи слово тяготеет к семантическому тождеству, в художественном тексте полисемантическое слово не только способно сохранять свою многозначность, но и стремится реализовать это свойство. Такая особенность функционирования многозначной лексики в художественном тексте обуславливается двумя определяющими характеристиками художественного текста -«стремлением к максимальной информационной насыщенности» (Ю. М. Лотман) и художественно-образной речевой конкретизацией. Повышенная смысловая нагрузка и образность оказываются непосредственными факторами эстетической (собственно художественной) ценности текста.Цель работы -в свете данных положений проанализировать семантико-стилистические особенности функционирования многозначного слова в поэзии А. А. Вознесенского, для творческой манеры которого характерны повышенная метафоричность, экспрессивность, неожиданные смысловые эффекты.Материал и методы. Основными методами исследования соответственно становятся семантико-стилистический анализ, контекстуальный анализ, эстетическая интерпретация.Основные результаты. С привлечением большого количества примеров показано, что в стихотворениях Вознесенского семантическое приращение и зримость образа достигаются как путем противопоставления значений многозначного слова, так и посредством их совмещения. Прослежено, как ресурс лексической многозначности искусно используется автором для усиления лиризма или драматизма текста, выражения авангардистской эстетики, философии взаимного уподобления и взаимосвязи всего сущего. Среди актуализированных автором приемов эстетического преломления многозначности слова оказываются семантические наслоения и сдвиги, акцентирование буквального значения слова, выдвижение одних смысловых компонентов слова и нейтрализация других, создание окказионального смысла слова в результате индивидуально-авторского ассоциативного переноса значения, прием «мерцания», неожиданная активизация имплицитного значения слова. Проиллюстрированы примерами приемы оживления стертых метафор и метонимий. Обозначены традиции В. В. Маяковского и Б. Л. Пастернака в некоторых особенностях метафорического переноса.Заключение. Результаты исследования могут быть интересны для уточнения особенностей идиостиля А. А. Вознесенского, стилистических и риторических функций полисемии, семантического потенциала слова.Ключевые слова: многозначное слово, семантика, стилистика текста, язык художественного текста, поэтическая речь, А. А. Вознесенский. А. Ю. Саркисова. Семантико-стилистические особенности...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2025 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.