How to Scrape and Visualize Weibo Comments with Python: A Step‑by‑Step Guide

Background – On September 21, Xue Zhi Qian posted a detailed rebuttal on Weibo to allegations made by Li Yutong, summarizing twelve key points of his side of the story.

Data Collection – Using Python, the author crawled the comments on Xue's reply, gathering 16,996 entries within an hour.

Importing the Data

After loading the dataset, the volume of comments allows a statistical view of netizen attitudes.

Word Segmentation

Code Example

The script splits each comment into words, producing a two‑column table: comment ID and segmented terms.

Removing Stop Words

Common Chinese stop words (e.g., “的”, “吧”, “这”) are filtered out before visualization.

Generating Word Clouds

The resulting word cloud shows dominant terms such as “support”, “believe”, and “evidence”, indicating that netizens largely trusted Xue’s explanation.

Comparative Analysis

Separate word clouds were created for comments posted before Xue’s reply and after it. Before the reply, the cloud highlighted neutral or skeptical terms; after the reply, words like “support”, “evidence”, and “face‑palm” appeared, reflecting a shift toward backing Xue.

This case demonstrates how Python can be applied to social‑media data mining, sentiment tracking, and visual storytelling.