Discover the users who are most similar to you based on your commenting style
How It Works
Data Collection: iFunny comments are collected from featured posts.
Semantic Embeddings: Using magic AI bullshit, recent comments are converted into high-dimensional vectors that capture their semantic meaning, context, and style.
User Profiling: Comment embeddings are aggregated to create a unique semantic profile for each user.
Cosine Similarity: When you search for a username, magic math bullshit is used to compare your semantic profile against all other users, finding those that match you best.
Frequently Asked Questions
What makes two users "similar"?
Top matches are determined by the cosine similarity scores between users' semantic profiles, as defined by the average semantic embeddings of their most recent comments.
How accurate or meaningful are the similarity scores?
Idk.
Is my data private and secure?
In general? No.
In the context of this site? Also no. This is just aggregation of publicly available information.
Why doesn't my username appear in results?
Your username might not appear if:
(1) You haven't made enough public comments recently for analysis
(2) Your comments were posted after the last data collection
How often is the database updated?
Whenever I feel like it or get around to it. I have to rerun a script each time that takes a few hours.
If I haven't done it in a while, send me an email.