![]() |
I wrote a machine learning tube indexer but im clueless about the industry
Hi I have done search algorithms and content recommendation systems at major tech companies and I wrote a tube indexer that does machine learned recommendations and has a good search engine. I have multiple clones of it live right now but I do not know how to market or seo them because I am completely new to porn and webmastering in general. Also the front ends suck cuz I am a backend guy. If you are interested in helping me monetize the work I have done I will pay you in cash or equity (I own an incorporated company under which I am conducting business). Basically just look at betterfap.com, their algorithms suck dick compared to mine but they are good at marketing and front end so they are popular and I am not. Can you help with that?
|
Welcome to GFY, can we see your site?
|
Welcome.
|
Sure why not.
pornharmony.com is currently getting the most traffic but I have a feeling I have already trashed it with spam blasts. It had a big spike in traffic from google that is dropping off rapidly. Oops. If you click on a few vids and return to the front page you get recommendations like what you've clicked on. If this means anything to you, its almost entirely content-based filtering at this point because collaborative filtering doesn't really do shit when you have 11 million videos and only a couple hundred thousand views of those videos. Its the catch 22 of ml; you need users for it to be really good, but it needs to be good to get users. Betterfap got around that with good marketing and nice design but truthfully even with all the data they have they still don't seem to be able to pick up on my fetishes, which my site definitely does. Tmi but I think they are doing latent feature discovery because that's the hot technique these days but anyone who's worked in ecommerce could tell you how pointless that is with a corpus as diverse as porn. |
I've been going through Bishop's book (Pattern Recognition and Machine Learning) and was wondering about the applications for adult as well.
I'm curious, what kind of model are you using for your tube example? Not sure how much information is available for tube sites since they have very short titles and maybe some tags is about all the information you can get about it. I'm thinking playlists on something like pornhub can be mined for additional information. The nature of playlists lends naturally to tight clustering. Are any adult companies publishing datasets like the "Netflix Prize" or the peer to peer lending datasets? That would be fun to analyze and play around with. |
Quote:
Dude.. Let's talk. I have a solid business plan. we've got Front end engineers and backend as well. My skype ID in sig. Ping me. |
betterfap.com what script is that exactly? Where can i throw my wallet at to get it?
|
Quote:
|
Nice site man, good luck! :thumbsup
|
Quote:
|
I can help with the optimization. Skype me
|
welcome buddy!
|
Quote:
I've got this for you. |
All times are GMT -7. The time now is 09:36 PM. |
Powered by vBulletin® Version 3.8.8
Copyright ©2000 - 2025, vBulletin Solutions, Inc.
©2000-, AI Media Network Inc