In addition to OSINT Tool Tuesday, I'm going to start doing Workflow Wednesday where I unpack a process, instead of a tool, for open source intelligence. This week I'm going to talk about how to deconstruct a new social media platform.
👇 (1/9)
Step 1: Map the platform without an account.
You want to see what you can access without registering. Explore the platform from the website but also check out what's indexed by Google and other search engines using site:, -site:, inurl:, intext:, and other operators.
👇 (2/9)
Step 2: Understand the platform's privacy policies and other fine print.
You want to see what the risks are for registering an account including what information is collected and shared. You also want to know what other users can view once you've registered.
👇 (3/9)
Step 3: Get the right tools
You want to find platform-specific tools or general tools for things like data collection, image capture, MHTML capture, video capture, metadata capture, etc. This can be a script or a browser extension.
👇 (4/9)
Step 4: Register an account and pay attention
Register an account using a sock puppet and see what happens immediately after. This includes what the default privacy settings are, what information like usernames are created by default, any follow up emails you get, etc.
👇 (5/9)
Step 5: Understand how the platform works
You want to understand what features the platform has and how the top influencers are using them. This will let you know which data points to prioritize. Additionally, see what other platforms are integrated/linked.
👇 (6/9)
Step 6: Study the Platform using DevTools
Open your DevTools and inspect the source code, what's going on in the Console, as well as the Network tab. This will let you know what's happening behind the scenes. Look for patterns in classes and ids as well as anomalies.
👇 (7/9)
Step 7: Extract a Sample Dataset
Use an extension (or script) like Instant Data Scraper to extract a sample dataset from a profile page and analyze the results in a spreadsheet. Use data visualization, link analysis, and formulas/macros to gain insights and find gaps.
👇 (8/9)
Step 8: Share your Findings
You probably overlooked something and sharing will help find out your gaps, create a feedback loop, and improve your process.
On that note, which steps did I miss?
(9/9)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
[#OSINT] You can use Twint to find indirect relationships between users. By matching the “conversation_id” to multiple queries, you can discover more insight.
For example, let’s say you’re trying to find violent users on Twitter that are threatening an influencer. Twitter only limits you to search for “influencer name” + “violent keyword”. Using Twint, you can search for all “violent keywords” and then match it to influencer mentions
You can also mine replies within Twint. It’s not in the Wiki, but adding “c.To =“ to your python module will allow you to pull tweets sent “to” someone. By finding the accounts that most mention said influencer, you can take this a a step further