Improve your masturbation efficiency!
What is it?
Porn Sieve learns your adult interests and then fetches videos that match those interests from xvideos.com. I am in no way associated with that website and in the future I plan on adding more websites.
Dependencies:
numpy
scipy
sklearn
pyside
How to use:
Run main.py with python3.
The top-left corner includes a list of fetishes to select.
Enter the page to start on and the number of pages to go through.
Press scrape when you want the program to load up porn for you. A
preview should appear on the screen if everything is working okay.
Recalculate predictions should be working properly now and redo
its prediction for the videos sitting in the video queue. This may
be necessary in the beginning when the program is still learning
your interests. Personally, I've found that it may be just as
important to rate videos you don't like as those you do. I try to
get an even split in the beginning before the thing warms up.
On the right is a slider, the rate button, the open button and the
skip button. When the slider is at the bottom, that inidicates a
very poor rating. Moving the nub all the way up to the top tells
the program this video really gets you.
DO NOT press rate before you hit open. Pressing open will open the
link in a browser window, whereas rate will record your vote and
move on. While the program does remember previous ratings, as of
yet there is not graphical way to get that data out of the database.
And at any rate, that'd probably be annoying.
Skip will move on to the next video, while failing to record a rating.
Pressing quit directly will save your model and encoder in its current
state. Without it, a new model must be train from scratch which depend-
ing on the amount of data saved, can take a long time.
Keybindings: o open current link in browser tab skip right skip left unskip (go back to previous) 0-9 set slider position enter rate video
Troubleshooting:
If anything goes wrong try:
1. Turning it on and then off again.
2. Deleting *.pkl files and then hitting "recalculate prediction model."
3. Deleting default.db (will remove your rate history).
4. Leave an issue so I can fix it.
Contributing:
Don't be shy about making changes or suggestions or asking me directly what
the hell I was thinking with a certain bit of code. Check out the git wiki
for a general overview of how I've organized the code.