Well, the problem is that my site is supposed to do exactly what reddit does. In fact, it was inspired by complains on reddit that an influx of new users buried the old interesting content. You do bring up a good point, however. I'll think more about differentiating!
My first impression of the site is that no big value jumps out at me when I look at it -- nothing makes me think, "Hey, this is really a lot better than trawling Google News and the Yahoo! Most Popular page, plus a few niche sites I find interesting."
Also, "Interesting things to read" is a little ambiguous, and could use clarification to emphasize that the site _does_ cover both news and reference texts. (Or is it even more than that?) With so much flat-looking text, people are going to read one link and think, "This is just another less-attractive version of Google News," or, "This is just another link aggregator site, but without arrows for some reason."
Somehow, you've got to make your front-page impression for non-logged-in users both provide a little bit of value (but not the many links you're putting up right now), _and_ immediately and visually demonstrate what makes your site uniquely valuable.
The list format feels a little oppressive, for some reason. I can't put my finger on it, but it just isn't ... fun.
Finally, all the links I see initially are about equally interesting to me -- they're all fairly heady math/computer geek stuff. So I don't even get the initial reward of trimming away (or the fun of condemning) the really trivial stuff I wish wasn't in the news. I wonder if someone in the 99% of the U.S. population that doesn't care about Boltzmann machines would even try to push past that.
I signed up for your site a while ago so I'd like to share my thoughts.
Design wise the site is a bit dull, but passable. The text is unbelievably hard to read (on Windows/Firefox, default font size). I'm guessing you use Mac or Linux, otherwise you would never have launched with that font.
I originally signed up because the site had the promise of a good recommendation engine. That's something that others have promised (Jaanix, reddit) but no one has really delivered on. If you have got something that you think is unique then tell people about it. I even visited the about page on my first session, trying to find what made your page different. I still don't know.
The biggest problem was with the content. When I first visited most links on the front page seemed to be on machine learning/AI. I guess that probably is because of the interests of the initial users, but I don't need a whole page of links on the same topic. If possible a mix of article types would be good; I found most links were just Wikipedia articles.
Finally, I'm not sure how I feel about the rating system. Other sites have two ratings for a reason. It is much harder to choose between 10 options and breaks up the flow.
I remember that and I think it is interesting. I have bookmarked it, but somehow I did not check back often. For one thing, I am busy keeping up with Hacker News. But maybe there are other things at work - perhaps you just need to improve the design a bit, and give it more of a community feel (ie make it clear what it is all about)?
A related question: does anyone know how reddit/digg/facebook/myspace gained their initial traction? Facebook couldn't have been very useful with all of 5 people on it. How did they gain their initial group of users?
I was one of the first couple hundred people on Facebook (and one of the first 30 people at MIT on it). Zuckerberg started it at Harvard. It spread like wildfire at Harvard because it was highly useful and then he opened it up to MIT. And then Yale. And then UPenn. And so on and so forth.
I think the takeaway there is to try niching to a manageable segment of the user population at first.
“I was one of the first couple hundred people on Facebook (and one of the first 30 people at MIT on it).”
I hate to rain on your parade (and this is admittedly extremely nit-picky), but those two statements are incompatible, unless you had email addresses from both institutions. Something like ten thousand students from Harvard, Columbia, Stanford, and Yale had joined the site before MIT was added.
Hmm. If you're right, I feel like a bit of a doofus. That's not as I remember it, but I don't have evidence to back up my point except my (probably flawed) memory. I will, however, stand by the weaker statement of being one of the first few from MIT.
And don't worry, that's not much of a parade to rain on. ;)
reddit: PG and Joel Spolsky talked about it.
facebook: Zuck acquired the main harvard email list and spammed everyone.
myspace: The parent company spammed all its users.
digg: ?
I think spamming is almost inevitable. The plentyoffish guy spammed myspace (i think) and made lots of female sockpuppets.
From what I remember, Digg first became popular when Paris Hilton's phone got hacked.
They were one of the first sites to break the news with the leaked phone numbers.
True. We were chugging along, growing about 20% a month, 4 months in, when that story appeared on digg and then got indexed by yahoo (we were no.1 and no.3 for the search "paris hilton phone"). Traffic doubled instantly. That should underline the importance of SEO. Get your friends to link to your site.
Yeah, I tried to do some reading about plentyoffish and couldn't find anything about myspace spamming. Though I don't doubt he had a ton of fake accounts at the start, and probably did some deal of spam.
Facebook gained its initial user base by recruiting a huge percentage of the Harvard student body:
“Today, two weeks after its inception, thefacebook.com has blossomed from about 650 members early last week to a network of over 4,300 student, alumni and faculty subscribers as of yesterday.”
According to interviews the Reddits have given, they (& friends & fellow YC founders) basically submitted everything themselves for the first 2-3 months. Any time they ran across something interesting on the net, they threw it on Reddit.
I remember checking it out the day it launched and leaving because it wasn't very interesting. I came back in October (about 3-4 months later) and it had some minimal traction. IMHO, the critical factor was that they added comments and people had started commenting on links.
Facebook was pretty well-established when I joined (fall 04, about 8 months after it started). People were using it as an address book - meet someone at a party, you instantly had all their contact info once you knew their name. And since there were pictures and it was organized by college, you could often pick out who they were from first names alone. There weren't many social features back then; IIRC, it just had Poke and the Wall.
I have a plan to start a game of some sort, get a local computer store to sponsor it with a new PC as the prize.
Maybe sell local advertising on the site as much to spread awearness as anything.
If it goes well then find a bigger sponsor and do another game, slightly less local next time around....
Congratulations, I have written one too, what a coincidence (sarcasm directed towards situation, not the person). I bet most of the people reading this thread have done something around netflix results :)
I have looked into SVMs but I don't think they would work well in this case because:
1) A separate classifier would have to trained for each user and this would take too much resources.
2) I think an SVM would require too many training cases before it becomes useful.
This is a personalized news site that I wrote in two months based on algorithms from the book Programming Collective Intelligence. Please tell me what you think.
It has two main features: the ability to identify related /similar links and suggestions/recommendations that actually work.
The basis for all of the algorithms is a document similarity metric presented in Chapter 3: Discovering Groups. Basically, to compare document A with document B, we calculate the Pearson correlation coefficient between the word frequencies of document A and the word counts of document B. (You can imagine this as plotting a series of points of a graph: each point's x coordinate is its frequency in document A and each point's Y coordinate is its frequency in document B. The Pearson correlation coefficient is a measure of how well the line-of-best-fit fits the points.)
Using this similarity metric, links can be clustered together using K-means clustering. This is what you get when you click on “related” at the bottom of each link. Clicking on “similar” gives the results of running K-NN. (“related” doesn't work as well as it could be right now because there are too few links for a link to be similar with, but this is an example of where it does work: http://fyynd.com/links/197/related/ “similar” usually works better right now.)
There are two algorithms for giving recommendations, “Suggested” and “Recommended”. "Recommended" generally works better than Suggested when you haven't yet made votes but Suggested should be more in tune to your preferences in the long run.
In layman's terms, the Recommendation algorithm works by "averaging" together the links that you liked and then find links that are similar to that while the Suggestion algorithm tries to determine whether you will like a particular link by seeing whether it is similar to any page that you have already rated highly. As a result, "Recommended" will list pages in your general interest area, but insensitive to any "niche" interest that you might have. The "Suggested" page will be sensitive to "niche" interests but will requires more votes to train. For example, if most of the link you rate highly are about computer science, with a only a few links about biology, when the recommendation algorithm averages them together, the biology links would count for very little. As a result, you wouldn't see much on biology. On the other hand, the suggestion algorithm will not be hindered by this, though it will have trouble if you don't vote much.
Please note that because predictions are so computationally intensive, they are not updated in real-time but on a hourly basis. Thus, you have to wait a bit before they come out. Please be patient!
Please check it out and tell me what you think! Any questions/comments/suggestions are more than welcome!
P.S.: I forgot to mention: the voting system normalizes your ratings. Thus, if you vote all 5 stars it is same as not voting at all! You must tell it what you don't like as well as what you like.
I really like the interface. It has some features I wish HN had, like the ability to hide items from view. I've been meaning to write something like this for a while but never got around to it. Keep up the good work.
oh, please create a bookmarklet to let users submit stories while browsing, this is VERY IMPORTANT, and shouldn't take much effort (use the HN one as an example).
I'd like to feed the site with stories from here and create a Greasemonkey plugin to automatically rate items on your site when I vote them up here (if I can find a good way to vote up items programatically on fyynd).
Bookmarklets: done. See http://fyynd.com/bookmarklets/. As for rating links programmatically: it is a simple POST to "http://fyynd.com/links/[link_id]/rate/" with a parameter "rating". "rating" should be a float between 0 and 5. A rating of 0 will delete that vote.
