OPORA continues to develop and publish our technical solutions to collect and analyze data from social media.
Our objective is to make the monitoring process more understandable and accessible and to automate most of its stages. We try to ensure that anyone, even without a special technical background, can use the data processing tools. Previously, we have already explained how to analyze the ads on social media Meta (Facebook, Instagram, etc.).
Now, we publish our code you can use to check how similar the posts in different Telegram channels or Twitter accounts are. For example, if you feel like a post on a Telegram channel is suspiciously similar to a post on another channel, the code can help you verify how often other posts from those channels overlap. The code could be useful for researchers looking for efficient ways to process big data sets.
CODE
In April, OPORA published a research based on the application of this code. It helped us find out that 53% of posts from pro-Russian Telegram channels in Ukraine’s temporarily occupied territories are not genuine but merely copy-paste each other.
The code is available in the Google Colab and includes additional user guidelines inside. To use the code, copy this text box into your workspace: “File” → “Save copy on the Disc”.
Should you have any questions, feel free to contact via our official email [email protected] or directly to the research team ([email protected], [email protected]).