Journal of Language Situation and Language Service
JLSLS, Vol. 1, No.1, 2026, pp.57-70
Print ISSN: 3078-3755; Online ISSN: 3104-5065
Journal homepage: https://www.lslsjournal.com
DOI：https://doi.org/10.64058/jlsls260105qsdcw
侯仁魁(Hou Renkui)，张锦程(Zhang Jincheng)
摘要：本文收集1993-2024年广州市政府工作报告建立语料库,选择表示主题的实词，包括名词和动词，基于向量空间模型建立文本向量，以文本聚类和对应分析为手段，考察报告中三十年来表示社会主题的名词和动词的发展变化。文本聚类结果显示，政府工作报告大致分为五个阶段，分别是1993-2000，2001-2008，2009-2012，2013-2019和2020-2024。从对应分析结果图中能够看出有些词语在某个阶段的报告中出现频率较高。接着以前一阶段报告为参照语料库（Reference Corpus），计算当前阶段关键词，关键词或者是在当前阶段报告中第一次出现，或者出现频率大幅提高，更能明确地观察到词语的变化，从而了解报告专注点的变化。最后，采用计量语言学方法抽取每个阶段报告文本的主题词（thematic word），计算它们的交集，从而发现所有阶段恒定不变的关注点，如“发展”、“经济”、“教育”、“改革”和“文化”等。
关键词：广州市政府工作报告，文本挖掘，主题，历时演变
作者简介：侯仁魁，通讯作者,副教授，广州大学人文学院、国家语言服务与粤港澳大湾区语言研究中心，应用语言学，电邮:hourk0917@163.com；张锦程：研究生，广州大学人文学院，语言学及应用语言学，电邮：2112412080@e.gzhu.edu.cn。
Title：Quantitative Study on Diachronic Changes of Words in Work Reports of Guangzhou Municipal Government
Abstract: This paper collected 32 Work Reports of Guangzhou Municipal Government from 1993 to 2024, to establish the corpus. The themes of the report were represented by the nouns
and verbs which represent the concrete and abstract objects and taken actions. Text clustering and correspondence analysis were used to explore the diachronic changes of themes represented by nouns and verbs. The results showed that Work Reports were divided into five phases which are 1993-2000, 2001-2008, 2009-2012, 2013-2019 and 2020-2024 respectively. The result of correspondence analysis demonstrated that some content words occur more frequently in some reports than in others. The keywords of Work Reports at each phase showed the changes of its focus. Lastly, we extract the thematic words of Work Report at each phase from the perspective of quantitative linguistics. The intersection of thematic words at each phases reveals the consistent and stable focuses of report over years, such as “发展(development and develop)”, “经济(economy)”, “教育(education)”, “改革(reform)” and “文化(culture)”. This paper explored the development and stability of focused themes of Work Reports of Guangzhou Municipal Government over 32 years.
Keywords: Work Report of Guangzhou Municipal Government, Text Mining, Thematic words, Diachronic change
Author Biographies: Renkui Hou, corresponding author, Associate Professor, School of Humanities, Guangzhou University; National Research Center for Language Service and Languages of the GuangDong-Hong Kong-Macao Greater Bay Area, Guangzhou, China; Applied Linguistics, E-mail: hourk0917@163.com. Jincheng Zhang, Master, School of Humanities, Guangzhou University, Guangzhou, China, Linguistics and applied linguistics, E-mail: 2112412080@e.gzhu.edu.cn.
Received: 01 Feb 2026 / Revised: 11 Mar 2026 / Accepted: 13 May 2026 / Published online: 30 May 2026 / Print published: 30 Jun 2026
