AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS最新文献_第4页

Hierarchical Methods for Automatic Classification of Scientific and Technical Documents according to the GRNTI Rubricator 基于GRNTI Rubricator的科技文献自动分类层次方法

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700512

I. A. Chechulin, Al. A. Khoroshilov, I. A. Danilov, M. A. Ivanov

{"title":"Hierarchical Methods for Automatic Classification of Scientific and Technical Documents according to the GRNTI Rubricator","authors":"I. A. Chechulin, Al. A. Khoroshilov, I. A. Danilov, M. A. Ivanov","doi":"10.3103/S0005105526700512","DOIUrl":"10.3103/S0005105526700512","url":null,"abstract":"Given the exponential growth in the volume of scientific and technical information, the task of automatic document classification is becoming critically important. The State Rubricator of Scientific and Technical Information (GRNTI), which features a three-level hierarchical structure, serves as a key tool for knowledge systematization in Russia and the CIS countries. Manual assignment of rubricator codes is labor-intensive and subjective, necessitating development of effective automatic methods that account for specificity of scientific language, class imbalance, and multilabel classification. The article examines approaches for automatic classification of Russian-language scientific and technical documents according to the GRNTI Rubricator. The study analyzes hierarchical methods for multiclass, multilabel classification based on text vectorization neural network models for training classifier models. An original algorithm for calculating significance weights of document keywords is proposed, along with the application of Manifold Mixup regularization and the focal loss function. The results of training classification models for scientific and technical texts according to GRNTI rubrics up to the third level inclusive are presented. Experiments conducted on a dataset of 2.2 million documents from VINITI RAS demonstrate that the best results are achieved using BERTA vectorization model and MLP classifier, yielding a weighted F1-score of 0.852 at the first level of GRNTI, 0.753 at the second level, and 0.552 at the third level.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"201 - 210"},"PeriodicalIF":0.4,"publicationDate":"2026-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149011476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Development of an Intelligent System for Forecasting the Dynamics of Changes in the State of the Earth’s Surface Using the Amur River As an Example 以阿穆尔河为例的地表变化动态预测智能系统的开发

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700445

I. N. Alkhimenko, Ya. Yu. Grigoryev, A. L. Grigoryeva

引用次数: 0

Development of a Machine Identity Management System for Microservices with a NoSQL Backend in Zero Trust Conditions 零信任条件下NoSQL后端微服务机器身份管理系统的开发

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700482

H. M. Kunniev

引用次数: 0

AI Copilot for the Design of a Radiation Protection Shield 人工智能副驾驶防辐射防护罩设计

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700561

L. A. Zinchenko, A. M. Chernetsov, V. V. Kazakov, E. S. Polyakov, E. N. Komkova, V. M. Kiseleva

引用次数: 0

Generating Temporal Signals from Static Images for Spiking Neural Networks 基于脉冲神经网络的静态图像时域信号生成

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700718

A. S. Toshchev

{"title":"Generating Temporal Signals from Static Images for Spiking Neural Networks","authors":"A. S. Toshchev","doi":"10.3103/S0005105526700718","DOIUrl":"10.3103/S0005105526700718","url":null,"abstract":"Spiking neural networks (SNNs), i.e., neural architectures that represent and transmit information in the form of temporally distributed spikes, require time-dependent input, whereas data in computer vision are most commonly available as static images. This study addresses the transformation pipeline “image → temporal signal → spikes” and examines how the choice of input encoding influences SNN training dynamics, spike activity density, and computational cost. The experimental section implements and compares two encoding families: first-spike-time encoding (Latency) and intensity-based Poisson encoding (Poisson). Within these families, four operating modes are considered: baseline Latency without background suppression, modified Latency with a silence threshold, stochastic Poisson, and deterministic Poisson. The evaluation employs the following metrics: the average number of spikes per sample, the number of synaptic operations, an energy-related proxy metric, and indicators characterizing competition among hidden-layer neurons. Experiments conducted on the MNIST dataset (60 000 training and 10 000 test images) using a network with a hidden layer of 100 neurons and a simulation horizon of 200 time steps demonstrate that all examined modes support stable learning without activity collapse. Among them, the modified Latency mode with a silence threshold of ({{x}_{{{text{min}}}}}) = 0.05 achieves the most favorable balance between useful activity and computational cost: at 323.41 spikes per sample, it requires 14925.09 synaptic operations, whereas the baseline Latency mode without background filtering, despite exhibiting a comparable level of output activity (311.22 spikes per sample), requires 78 400 synaptic operations.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"S403 - S408"},"PeriodicalIF":0.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Representation of Intraword Syntagmatic Relations in Vector Language Models 向量语言模型中词内句法关系的表示

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700640

D. K. Rodionova, O. A. Mitrofanova

{"title":"Representation of Intraword Syntagmatic Relations in Vector Language Models","authors":"D. K. Rodionova, O. A. Mitrofanova","doi":"10.3103/S0005105526700640","DOIUrl":"10.3103/S0005105526700640","url":null,"abstract":"The paper discusses semantic structure representation of derivatives in language models, taking into account the intraword syntagmatic relations between derivational morphemes. Experiments were conducted using morphemic models developed by the Russian National Corpus (RNC), as well as fastText and ruRoBERTa models. The study has verified the hypothesis on compositionality of derived words which are represented as aggregated morpheme vectors. In experiments we explore the representation of semantic relationships using fastText morpheme vectors and standard subword vectors in ruRoBERTa. The results indicate moderate sensitivity of fastText vectors to syntagmatic relations between morphemes as well as to derivational types. It was also found that aggregating morpheme vectors in fastText provides a better representation of semantic relations between words compared to aggregating subword vectors in ruRoBERTa. Standard BPE (Byte-Pair Encoding) and WordPiece tokenizers used in Transformer-based models are poorly interpretable with respect to linguistic data, as word segments do not always correspond to morphemes. The research problem lies in the need to assess the extent to which modern language models can capture linguistic attributes characterizing the relationships of derived words within word-formation families. The aim of the study is to evaluate the ability of predictive distributed vector embedding models to reproduce syntagmatic connections between morphemes within derived words and at the level of word-formation families in the Russian language. The obtained results encourage the development of neural network architectures that take into account syntagmatic relations between morphemes, the improvement of morpheme tokenizers, and their integration into language models.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"S353 - S359"},"PeriodicalIF":0.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Development of an Intelligent Search System for the Mathematical Archive of Publications 数学文献出版物智能检索系统的开发

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700627

A. A. Nasibulin, O. M. Ataeva

{"title":"Development of an Intelligent Search System for the Mathematical Archive of Publications","authors":"A. A. Nasibulin, O. M. Ataeva","doi":"10.3103/S0005105526700627","DOIUrl":"10.3103/S0005105526700627","url":null,"abstract":"A study was conducted on searching for similar documents. The goal was to create a recommendation algorithm for finding similar scientific articles in mathematics using a prioritized search of mathematical formulas with textual support. The text was converted from graphical to textual representation using OCR technology for subsequent analysis and indexing. During the analysis process, the text was divided into blocks, followed by the extraction of significant formulas, keywords, and phrases from the text. During the indexing process, a vector database was formed based on vector representations of formulas obtained through the embedding process. The indexing results were used to search for articles that are similar to the document submitted by the user to the algorithm input. A list of similar articles is displayed with results sorted by the metric of closeness of vector representations of formulas. The source data consisted of approximately 5000 scientific articles devoted to various studies on mathematical topics and presented as PDF files. The experiment was conducted based on data from specific library system content, but the proposed technology can be extended to other library systems, including those containing articles on other topics, such as physics and other exact sciences.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"S338 - S344"},"PeriodicalIF":0.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improving Short Text Classification Robustness to Stochastic Noise Based on Density-Driven Training Data Cleaning 基于密度驱动的训练数据清洗提高短文本分类对随机噪声的鲁棒性

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700536

B. B. Baishev, A. P. Khalov

引用次数: 0

The System for the Automatic Generation, Processing, and Management of Document Metadata in Digital Collections 数字馆藏文献元数据自动生成、处理与管理系统

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700664

A. R. Khamedzhanov

{"title":"The System for the Automatic Generation, Processing, and Management of Document Metadata in Digital Collections","authors":"A. R. Khamedzhanov","doi":"10.3103/S0005105526700664","DOIUrl":"10.3103/S0005105526700664","url":null,"abstract":"The publishing cycle is currently undergoing significant technological changes: automated publication management systems are being implemented, neural network technologies are being used for content processing, and tools for the intelligent analysis of scientific data are being actively developed. One of the key trends is the automation of the publishing cycle, aimed at accelerating manuscript processing, improving the quality of metadata, and ensuring the interoperability of information resources. In this context, metadata serves as a connecting element for machine processing and navigation within the scientific knowledge space, ensuring the structuring, interpretation, and integration of information into digital library systems. However, metadata for scientific publications often contain errors, inaccuracies, or are incomplete, and their manual creation and refinement are time-consuming and do not ensure high accuracy. The aim of this work is to design and develop a system for the automatic generation, processing, and management of metadata for scientific documents based on data obtained from scientific publication search services and open knowledge bases. The system can be used to automate the process of extracting, refining, and supplementing the metadata of scientific publications for the purpose of subsequently creating electronic collections of scientific documents.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"S367 - S375"},"PeriodicalIF":0.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Methods for the Automatic Assignment of UDC Codes to Mathematical Articles: Evaluation of Classical and Neural Approaches 数学文章自动分配UDC代码的方法：经典方法和神经方法的评价

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-11 DOI: 10.3103/S0005105526700548

B. T. Gizatullin, O. A. Nevzorova

{"title":"Methods for the Automatic Assignment of UDC Codes to Mathematical Articles: Evaluation of Classical and Neural Approaches","authors":"B. T. Gizatullin, O. A. Nevzorova","doi":"10.3103/S0005105526700548","DOIUrl":"10.3103/S0005105526700548","url":null,"abstract":"Universal Decimal Classification (UDC) is a hierarchical indexing system in which a publication can be assigned one or several codes. Manual UDC indexing is labor-intensive and often inconsistent. This paper examines the automatic assignment of UDC codes to Russian-language mathematical research articles. The aim is to compare combinations of text representations and classification models on a unified corpus and to identify the most effective configurations. A corpus of 4194 articles was collected from Math-Net.Ru, including full texts, abstracts, metadata, and UDC codes. The preprocessing pipeline comprised PDF text extraction, removal of layout artifacts, and normalization of UDC labels. We compared TF-IDF, Word2Vec, SciRus-tiny, and SciRus-tiny3.5 representations combined with logistic regression, Complement Naive Bayes (CNB), and CatBoost. In both the single-label and multi-label settings, the best performance was achieved by TF-IDF + LogReg, while TF-IDF + CNB showed closely competitive results. The proposed approach can be used in automatic subject indexing systems for digital libraries and scientific archives, in UDC recommendation tools for authors and editors, and in metadata quality control workflows.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"S277 - S284"},"PeriodicalIF":0.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148927745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0