Comment by noname123 | Hacker News Reader

Comment by noname123 | Hacker News Reader

This. As an avid porn user, I've been telling one of my friend who's in the business of recommendation engine to "pornify his business" for profit.

So much talent, data-mining has been spent on the problem of recommending what product you should also look at Amazon.com from the page you on, what Netflix film you should queue up next and what song you should listen to next according to the 512 acoustic attributes of the current one. Forgot all that, I want the next best clip to watch based on - one of the most frequently pursued and unacknowledged male (and female) activity on the Internet.

Data is all there. Tagging is categorized to the T on porntube sites by the actor/actress, genre, ethnicity and different acts depicted. Aggregator sites are already out there so all someone needs to do is build an aggregator site that plots out the user "interaction" graph from one clip to the next and make correlations. Next a snazzy Pandora interface where you could skip (unlimited times) to the next video, or show multiple preview of video's that you could skip to; or maybe make it social like turntable.fm where multiple people could vote to rock their socks off to a clip or not and see how long people in a particular room last... the possibilities are endless in pornification of the interwebz

> Data is all there. Tagging is categorized to the T on porntube sites by the actor/actress, genre, ethnicity and different acts depicted.

I looked into doing this.

The problem is that all of the popular tube sites have bad data especially in the tag section. People put up videos with the most searched on tags, even if the content doesn't match.

Actress/Actor data is a joke on many sites---being a wild guess at best especially for asian actresses.

Still you could take what they have and make it easier to search through. There's some movement on that front in the niches--like hentai, but none on the main stream sites.

One thing to worry about in recommendation engines (which is why they're so conservative on the big sites) is that users are enthusiastic about their particular niche but nothing else, and unlike in say Netflix--there's a real possibility of showing miscategorized guro to someone doing a vanilla search.

As a startup in the adult space I know there are lots of opportunities in adult...especially since the industry is going through a huge transition. As much as people love to say how porn brings innovation, that's not the case anymore. In the beginning it was way too easy for many of these porn sites to make money. Basically you just had to stick up a thumbnail gallery and a payment page and money would come in.

However, now as free porn is prolific on the internet many of these webmasters don't know what to do... so not that much innovation going on. If you can bring in good tech people there's no doubt they could kick butt. That's what we're trying to do, but just being in adult makes people look at you funny. Something that comes with the territory.

Reply on news.ycombinator.com