Tokenized Results of Comments on Bus Travel from Chinese Social Media Platforms

The data comes from Chinese social media platforms — MicroBlog, Redbook, and TikTok — and reflects public evaluations of various aspects of bus travel. The dataset used here is the result of preprocessing the raw data, which involved two main steps:
（1）Filtering Invalid Posts：To ensure data quality, invalid posts were removed.
Systematic Text Cleaning and Transformation
（2）The text was cleaned and transformed to facilitate subsequent analysis： This process included removing elements that might interfere with topic detection, performing tokenization and lemmatization to convert words to their base forms, and eliminating stopwords, excessively long words, and low-frequency words.
After preprocessing, a total of 6,403 valid comments related to bus travel were extracted from the three platforms, covering the period from January 2023 to December 2024.
Based on this dataset, seed word classification can be applied to assign each comment to a specific topic. After classification, TF-IDF can be performed on the dataset of a given topic to further extract the keywords from each comment, which can then serve as the foundation for topic modeling.