文献类型：期刊文献

英文题名：Research on Key Technology of Web Topic Detection Based on Tuple Semantic Description Analysis for Big Data

作者：Chen, Mo[1]

第一作者：陈默

通讯作者：Chen, M[1]

机构：[1]Beijing Union Univ, Sch Business, Big Data Management & Applicat Major, 3 Yanjing Dongli, Beijing, Peoples R China

第一机构：北京联合大学商务学院

通讯机构：[1]corresponding author), Beijing Union Univ, Sch Business, Big Data Management & Applicat Major, 3 Yanjing Dongli, Beijing, Peoples R China.|[1141721]北京联合大学商务学院;[11417]北京联合大学;

年份：2026

卷号：33

期号：1

起止页码：225-236

外文期刊名：TEHNICKI VJESNIK-TECHNICAL GAZETTE

收录：;WOS:【SCI-EXPANDED(收录号:WOS:001653669500025)】；

基金：This paper is supported by General Project of Science and Technology Plan of Beijing Municipal Education Commission under Grant Nos.KM202011417011, Research Project on Graduate Education Science at Beijing Union University in 2025 under Grant Nos.YK202502, Support Project of High-Level Teachers in Beijing Municipal Universities in the Period of 13th Five-Year Plan under Grant Nos.CIT&TCD201704072.

语种：英文

外文关键词：big data; tuple semantic description; web topic detection

摘要：In the context of the numerical intelligence age, how to further promote research on Web topic detection with heterogeneous big data as an important research object has attracted widespread attention from scholars. This paper takes network heterogeneous big data as the main research object and proposes a Web topic detection idea based on tuple semantic description analysis. Firstly, the main implementation methods are explained, including time item analysis, named entity item analysis, event item analysis, and semantic formal analysis. Secondly, the Web topic detection algorithm based on tuple semantic description analysis is elaborated. Through experiments, firstly, the impact of the quantity of Web news on the quality of time item description and five tuple semantic description is analysed, the influence of the adjustment parameters in the algorithm on the quality of the described Web news named entity items is also analysed to obtain the optimal adjustment range for these parameters, the impact of the long and short time selection range in the algorithm on the calculation results of the popularity weight of Web news named entity items is also analysed to obtain its optimal adjustment range, the impact of the adjustment parameter in the algorithm on the quality of the described Web news event items is also analysed to obtain the optimal adjustment range of the parameter. Secondly, the quality of topic detection under different semantic descriptions in Web news, the time consumed in the topic detection process under different methods, and the quality of topic detection under different datasets are analyzed. The experimental analysis process shows that the Web topic detection idea proposed in this paper is feasible, verifiable, and superior, and can play an important role in reconfiguring the Web topic corpus, inferring the Web hierarchical big data propagation path, and providing the numerical intelligence warehouses based on network information detection.