Popularity does not equate to relevance
I’m a big advocate of community driven content on news sites.
Having said that, I’ve recently been thinking that it’s not always best that the community drives everything.
Digg.com is a prime suspect.
Good site but it has its flaws. Namely, many of its users are not particularly good samples of human intellect.
Don’t believe me? Take a random sample of the stories and comments posted and see for yourself.
So, here’s the problem. Digg works on the theory that popular is relevant. The majority of our populous is not relevant to my interests, so why would I trust “most people” to choose content for me?
The trick lies in relevant people choosing relevant content.
I believe that in order for a community driven information source of any kind to offer value, they need to be able to map readers to authors.
Digg needs to somehow figure out what I like based on what I digg and give weight to stories that are not only popular but relevant to me.
That’s one of the fundamental principals of citizen journalism. You subscribe to authors that filter content based on your interests and speak in topics that you value.
Look at Engadget as a predominant example. It is driven by regular people who want to produce the best content and it so happens that they have become trusted to deliver everything gadget.
They still have a strong community aspect to them but don’t lose sight of the fact that it’s about quality, relevant content.
I have noticed that digg has recently added the “digging” capability to specific comments so that you can digg a comment and filter to the most popular comments on a post.
Nice bit of candy for sure, but like candy, there’s no real substance.
So, using the digg scenario a useful step forward might be this – Logged in under my profile, when I digg a story or commenter, the author goes into my digg list and extra weight is given to articles which they’ve commented on or submitted in the future. The front end is already there, they probably just need to put some smarts into the back end engine.
Until then, sites like digg will be nothing more than entertainment or boredom relief.
Now I know I’ve focused on digg here. The intention is not really to single out digg but rather to use it as an example.
TailRank has started to move in the right direction, they seem to have added a feature where you can filter by your own OPML file or by popular blogs, there’s that damn word again – popular!
Other sites such as tech.memeorandum, newsvine etc are also culprits. I’m constantly left wanting content relevant to me and not the result of mass popularity.
-dg
I’m no digg bitch, however…
First off – I take offence at your assertion that candy has no substance. I want to follow that up with either “she’s more than just a pretty face” or “you must be shopping at the wrong places” but can’t decide which way to go. ;-)
I agree that all “top” lists are flawed in as much as they are self-perpetuating. This manifests itself in best seller lists (yeah Dan Brown – I’m talking about you) in addition to the front page of digg. People who don’t have time (or the inclination) to look randomly, alphabetically, or chronologically, etc will gravitate to these lists and so all items appearing here have significantly more eyeballs that any other item – and as a consequence are more likely to be validated by purchasing or digging which reinforces their position on the list. I would say that digg addresses this better than most in as much as negative reviews can quickly demote an item off the list (front page).
I think your referencing “authors” is confusing. Are you talking about the article author or the digg submitter? To my mind the author is not relevant as if you like a particular author you are more likely to skip digg and frequent their sites. If you are talking about the submitter, I would say that they are only slightly relevant. The reason for this is if a story becomes available, there is a (may be a) flurry of activity to submit it to other channels (like digg). If a “trusted” submitter is too slow at submitting this article, it doesn’t make it less relevant.
For me, the mechanism for identifying the likelihood an article is relevant is the appraisal the actual article gets from a “trusted source”. In digg, this is someone in your “friends” list. Subscribing to this feed in preference to the front page will allow you to see a list of article with a higher level of relevance to you. If people submit articles (and by implication have dugg the article) or have simply dugg an article that resonates with you, you are then able to see other articles they have dugg and evaluate how close this person’s interests are aligned with yours. If you feel they share your opinion of “relevance” you can then add them as a friend and have them filter articles for you.
Now, in the spirit of community, if there are people who have you as a friend, you should digg articles that you feel are interesting or relevant so that you are sharing the load of filtering. You might also like to visit the front page or topic-specific pages periodically when you have the available time to see if there are articles that slipped through the cracks and be able to promote them to your “friends”.
I am a bit of a fan of this paradigm and believe that although it’s not perfect – it’s a great attempt at addressing the issues you raised.
phew!
p.s. I don’t digg shit so don’t waste your time adding me as a friend ;-)
Digg Bitch: A female dog that diggs?
Barbwire, many good points – all very good additional “hooks” for finding the elusive relevance gene (why didn’t I think of that term when I initially posted damn it!) in the mass of content available out there.
My comments . . .
You’re right, and I’m one of these people. And I suspect there are many others too. The whole point is that in order for these sites to be of value, they need to do some work for me and produce relevant content in a consolidated view. I can always look for content in many places, I have that already.
I’m referring to the digg submitter.
Very good point, the “friends” mechanism could also be used to deduce relevant, profile specific content. The issue with the digg implementation is that it does not contribute to the main page, or as you put it “the list”.
I know I’m harping on the main page. It’s significant because that’s why I go to a news these types of sites, to find a consolidated view of relevant news.
Again, this could be another good way to measure my interests. See what topic areas I visit most and weigh the main page based on that too.
On a final note, I believe that a superior profile specific relevance capability will be a major contributing factor as to what will separate the men from the boys in this space. It is what will make this next generation of community driven news a true force to be reckoned with and may give the biggest competitive advantage to whoever does it right.
-dg
Now would you look at that . . .
From Digg blog today:
-dg
Maybe what is needed here is a news genome project (in the same spirit of the music genome project) – something that scientifically and unemotionally categorises article content in accordance to various factors such as technology, style, currency etc.
Very nice. I like it.
Music Genome Project is manually populated. I would imagine that if the human/manual approach were taken, it’d be difficult considering the real-time nature of news.
An automated sequencing algorithm, now that would rock!
-dg