Doc2Vec-based similarity analysis of tobacco standard texts
Bin Li, Wei Wang, Zhuoyu Huang, Zhimei Liu, Hao Wang, Qingchun Feng, Na Zhao · International Journal of Modern Physics C · 2025
In the modern tobacco industry, standardized business processes are the core elements for ensuring product quality and enhancing production efficiency. The tobacco sector establishes strict standard texts to ensure that products meet requirements. At the same time, to accommodate product changes and improve efficiency, these tobacco standard documents are continuously updated over time. How to effectively find similar documents for reference during this process, thereby enhancing the standardization of documents and writing efficiency, is an urgent and complex problem that needs to be solved. This paper proposes a solution and a general framework — the Doc2Vec-based framework for similarity analysis of tobacco standard texts. Through experiments, it has been proven that this methodological framework demonstrates good performance, with high accuracy and reliability, and can quickly and accurately identify similar standard documents. The research in this paper not only promotes the development of text similarity analysis technology but also expands its application scope by applying it to tobacco business processes. In terms of practical application, the research in this paper will provide the tobacco industry with a good method and general framework to improve the quality of tobacco products, enhance production efficiency, and ensure that texts comply with industry standards and regulatory requirements. It also provides a reference for other industries facing similar issues.