Replication data for: A corpus-assisted sentiment analysis of Chinese social media discourse on the Israeli-Palestinian conflict

TitleReplication Package – A Corpus-Assisted Sentiment Analysis of Chinese Social Media Discourse on the Israeli-Palestinian Conflict: A Study of Machine-Translated Weibo Posts AuthorAnonymous DateDecember 2025 Version1.0 LicenseCreative Commons Attribution 4.0 International (CC-BY 4.0) DOIhttps://doi.org/10.5281/zenodo.18005944 DescriptionThis replication package contains all data, code, figures, and supplementary tables necessary to fully reproduce the results reported in the paper “A Corpus-Assisted Sentiment Analysis of Chinese Social Media Discourse on the Israeli-Palestinian Conflict: A Study of Machine-Translated Weibo Posts”. All user identifiers, usernames, profile links, and any potentially identifying information have been irreversibly removed. The package contains only publicly available, anonymised Weibo posts collected between October 2023 and January 2024. Contents├── data/ ├── weibo_cleaned.csv → Final cleaned corpus (8,046 unique posts) ├── posts_1000_translated.csv → 1001-post validation subsample with GPT-4 translations ├── posts_1000_translated_updated.xlsx → Same subsample with additional metadata ├── posts_1000_with_GLM4_FAST.xlsx → GLM-4-flash sentiment scores on subsample └── Manual_Coding_240_Posts_...xlsx → Inter-coder reliability sample (240 posts) ├── code/ → 18 Python scripts (.py) that perform cleaning, sentiment scoring, validation, figures, tables, and clustering (run in numerical order for full replication) ├── figures/ → All 11 publication-quality figures exactly as they appear in the manuscript ├── Appendix_tables.xlsx and .csv/ → Full tables for Appendices A–I (keywords, collocations, n-grams, 80 extreme translation cases, GLM-4 prompts, etc.) Software requirementsPython 3.10+ with the following packages:pandas, numpy, matplotlib, seaborn, scikit-learn, jieba, snownlp, vaderSentiment, plotdb### 复现包标题 基于语料库辅助的巴以冲突中文社交媒体话语情感分析：机器翻译微博帖子研究 复现包 ### 作者 匿名作者 ### 发布日期 2025年12月 ### 版本 1.0 ### 许可协议 知识共享署名4.0国际版（Creative Commons Attribution 4.0 International，CC-BY 4.0） ### 数字对象标识符 https://doi.org/10.5281/zenodo.18005944 ### 项目描述 本复现包包含复现论文《基于语料库辅助的巴以冲突中文社交媒体话语情感分析：机器翻译微博帖子研究》中报告的所有结果所需的全部数据、代码、图表及补充表格。 所有用户标识符、用户名、个人资料链接及任何可识别个人的信息均已被不可逆地移除。本包仅包含2023年10月至2024年1月期间收集的公开可用、已匿名化的微博帖子。 ### 内容结构 ├── data/ │ ├── weibo_cleaned.csv → 最终清洗后的语料库（8046条唯一帖子） │ ├── posts_1000_translated.csv → 包含1001条帖子的验证子样本，附GPT-4翻译结果 │ ├── posts_1000_translated_updated.xlsx → 同一子样本，新增元数据 │ ├── posts_1000_with_GLM4_FAST.xlsx → 子样本的GLM-4-flash情感评分结果 │ └── Manual_Coding_240_Posts_...xlsx → 编码器间信度检验样本（240条帖子） ├── code/ │ → 18个Python脚本（.py格式），用于执行数据清洗、情感评分、验证、图表生成、表格制作及聚类分析，需按数字顺序运行以完成完整复现 ├── figures/ │ → 与论文手稿中完全一致的11张符合出版标准的图表 ├── Appendix_tables.xlsx 和 .csv/ │ → 附录A至I的完整表格（涵盖关键词、词语搭配、n元语法、80个极端翻译案例、GLM-4提示词等） ### 软件需求 Python 3.10及以上版本，需安装以下依赖包：pandas、numpy、matplotlib、seaborn、scikit-learn、jieba、snownlp、vaderSentiment、plotdb