This repository contains code and analysis from the ODU 2023 REU project "Discovering the traces of disinformation on Instagram" by Rachel Zheng (@rachelzheng03).

The Python script takes the URI-M as a command line argument. For example, if you want to scrape information from https://web.archive.org/web/20170214033011/https://www.instagram.com/beyonce/, then you would navigate to the folder that contains the Python script and run the script with the following piece of code:

python instagram_memento_scraper.py --urim https://web.archive.org/web/20170214033011/https://www.instagram.com/beyonce/

The Instagram memento scraper outputs a JSON object, which can be interpreted as a Python dictionary. The output JSON object contains two main keys, "profileUser" and "userMedia." The value of the key "profileUser" is a dictionary that contains basic user profile information, including the username, biography, and follower count, on the Instagram handle associated with the URI-M. The value of the key "userMedia" is a list of dictionaries in which each dictionary represents a single Instagram post on the webpage associated with the URI-M. Each dictionary might include data such as the caption, comment count, or like count of an Instagram post. The amount and kind of information that can be scraped from an Instagram memento depends on the timestamp of the memento. For a detailed description of the information that can be scraped by my Python script, see the Table of Information Extractable from Mementos.pdf.

You can also direct the output to a JSON file from the command line. For instance, if you want to redirect the output from the example above to a JSON file named example14.json, you can put the following code in the command line:

python instagram_memento_scraper.py --urim https://web.archive.org/web/20170214033011/https://www.instagram.com/beyonce/ > example14.json

The full output for this example can be found at example14.json.

• Rachel Zheng and Michele C. Weigle, "Examining the Challenges in Archiving Instagram", Technical report arXiv:2401.02029, arXiv, January 2024, https://arxiv.org/abs/2401.02029.

This project was part of the 2023 ODU NSF REU Site in Disinformation Detection and Analytics.