AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS最新文献_第3页

Constrained Decoding for Structural Error Suppression in Dependency Tree Generation with Large Language Models 大型语言模型依赖树生成中结构错误抑制的约束解码

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-10-13 DOI: 10.3103/S0005105526700810

E. D. Shamaeva

引用次数: 0

Taxonomy Exploration with Large Language Model Reasoning and Tool Calling 基于大语言模型推理和工具调用的分类法探索

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-10-13 DOI: 10.3103/S0005105526700780

F. A. Sadkovskii, M. M. Tikhomirov, N. V. Loukachevitch

引用次数: 0

Application of Graph Neural Networks for Automated Verification of Building Information Models 图神经网络在建筑信息模型自动验证中的应用

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-10-13 DOI: 10.3103/S0005105526700871

O. V. Poletkina, O. M. Ataeva

{"title":"Application of Graph Neural Networks for Automated Verification of Building Information Models","authors":"O. V. Poletkina, O. M. Ataeva","doi":"10.3103/S0005105526700871","DOIUrl":"10.3103/S0005105526700871","url":null,"abstract":"Automating the verification of building information models for compliance with fire safety regulations remains an important task for the architectural and construction industry. Existing automated verification systems rely primarily on rule-based approaches that do not adequately account for the topological context of the building and are difficult to adapt to new projects.The paper proposes and experimentally evaluates methods for predicting fire-protection parameters of doors in building information models using graph neural networks and validates the approach on real design data from seven residential buildings provided by a major developer. A methodology for predicting the fire-resistance class of doors based on relational graph convolutional networks is proposed, and a pipeline for extracting data from a specialized format, constructing a graph, and generating features that account for geometric, semantic, and topological characteristics is developed. Experiments were conducted to predict whether fire-resistance requirements apply and, where applicable, the required fire-resistance class, using cross-project Leave-One-Building-Out validation. The developed approach makes it possible to automate the verification of fire-protection parameters and reduce the time required to analyze building models. The use of graph neural networks enables the model to account for topological context while achieving high prediction accuracy, and the use of real data confirms the practical applicability of the method.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 4","pages":"S523 - S530"},"PeriodicalIF":0.4,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176601","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 a Digital Intelligent Geological Knowledge Sphere 数字智能地质知识圈的开发

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-10-13 DOI: 10.3103/S000510552670086X

V. V. Naumova

引用次数: 0

Classifying Network of Database Subject Diagrams 数据库主题图分类网络

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-10-13 DOI: 10.3103/S0005105526700937

A. N. Rodionov

{"title":"Classifying Network of Database Subject Diagrams","authors":"A. N. Rodionov","doi":"10.3103/S0005105526700937","DOIUrl":"10.3103/S0005105526700937","url":null,"abstract":"In conceptual data modeling, the concepts of multilevel modeling and classification are inseparably connected with each other, since classification always underpins multilevel organization. Data structures specializing in placing classifiers can be found in the vast majority of modern databases. Classifying structures attached to object types simultaneously perform several important functions in databases as digital counterparts of reality. By means of semantic classifiers, which contain relevant for the concrete domain semantic sets, represented by nouns and adjectives, the primary semantic identification of objects is performed. Other classifiers focus on the physical characteristics of objects and the values assigned to these characteristics, thereby forming subsets of alike objects and additionally semantically identifying the latter. Classifiers, as a rule, form communities within which there is a clear trend for hierarchy. Being a product of human consciousness, hierarchical classifications are the most common form of classifier organization. Meanwhile, the horizontal, multivalued relationships that arise between classifiers are also an equally common reality. The structure template being developed come with universality and is able to locate in its structures an innumerable diversity of both existing and promising object classifiers. One of the distinctive features of the created model is the ability to automatically endow objects with relevant properties at the time of attaching classifiers to objects. This option is provided by switching to an invariant classification scheme, which is based on a questionnaire format for representing in modeling structures objects' properties.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 4","pages":"211 - 229"},"PeriodicalIF":0.4,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176713","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

Features of Collecting and Storing Multiformat Data in the Electronic Information and Educational Environment of the University 高校电子信息教育环境中多格式数据采集与存储的特点

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700469

A. R. Zhetev, V. M. Tyutyunnik

{"title":"Features of Collecting and Storing Multiformat Data in the Electronic Information and Educational Environment of the University","authors":"A. R. Zhetev, V. M. Tyutyunnik","doi":"10.3103/S0005105526700469","DOIUrl":"10.3103/S0005105526700469","url":null,"abstract":"To implement the project of the electronic information and educational environment (EIEE) of the university, the database structure, principles of database organization for working with multiformat data were developed, the role of multiformat databases in ensuring the functioning of EIEE and the characteristics of time series databases are shown. Methods of standardization and unification of heterogeneous data for subsequent processing and a pipeline of unified query processing in a multiformat EIEE storage are proposed. A model of integration of multiformat data and a model of information resources are presented. The principles of integration of multiformat data into EIEE have been developed. The proposed model of information resources makes it possible to classify and structure the entire variety of educational data, and the data integration model provides mechanisms for combining them into a single information space to ensure semantic consistency. The prospects for the development of this area in the distance learning system are related to the development of effective algorithms for the semantic display of data and the optimization of distributed computing in heterogeneous environments.","PeriodicalId":42995,"journal":{"name":"AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS","volume":"60 3","pages":"172 - 183"},"PeriodicalIF":0.4,"publicationDate":"2026-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149011475","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

On Certain Aspects of the Concept of Artificial General Intelligence 论通用人工智能概念的若干方面

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700494

S. V. Leshchev, K. A. Kalushev

引用次数: 0

Automated Extraction of Factual Information from Scientific and Technical Documents 从科学技术文件中自动提取事实信息

IF 0.4

AUTOMATIC DOCUMENTATION AND MATHEMATICAL LINGUISTICS Pub Date : 2026-09-22 DOI: 10.3103/S0005105526700500

A. V. Rykalin, A. A. Khoroshilov, S. E. Churilov

引用次数: 0

Hybrid Approach to Neural Network Training: Combining the Adam Method with Sparsifying Algorithms 神经网络训练的混合方法：结合Adam方法和稀疏化算法