-
Website
http://www.loiclemeur.com/ -
Original page
http://www.loiclemeur.com/english/2008/12/twitter-we-need-search-by-authority.html -
Subscribe
All Comments -
Community
-
-
Top Commenters
-
Popular Threads
All Comments
There are many more attributes that have higher value that are not available in Twitter search. I would love to just search messages from people in my network, but I can't do that just yet.
We need to keep EGO out of Twitter, or it will turn into another irrelevant network.
Apolinaras "Apollo" Sinkevicius
Twitter should let you search by number of retweets. Yes, I'd like to be able to see results in order of number of followers too. It's just that Twitter search sucks for a whole lot of reasons that have nothing to do with authority. Fix those first.
Literally stunned that we are still dealing with this elitist mindset days before 2009.
'Authority' isn't what's important, it's the strength of the connections you have, and the context of the endorsement. A 'I just love this blog, and this post is why' tweet from a Twitter user with 1,000 followers almost always trumps a tweet from someone with 10K followers, that puts no context to the link.
You want to communicate with ANYONE that's talking about your brand, not just the 'A-Listers'. I have lost track of how many people I follow on Twitter that only had a 100 or so followers a few months ago, that now have thousands. That's an INCREDIBLE expansion of influence! And because I didn't give a rip about their 'authority' and connected with them early on because they were smart and I could see that they 'got' the idea of contributing value to a larger community. They continue to tweet my links now as their community is expanding MUCH faster than the ones that some deem have 'authority'.
Did I mention that I am stunned that this is still actually being discussed in late 2008? You guys keep following the people you think have 'authority', I'll keep following the people that are actually changing this space.
The URLs are often shortened, so they provide no real value in searching. The tweets you are searching are often no longer than the keywords you are looking for.
If you are searching for "travel deals", you will have to deal with hundreds of tweets that contain little more information than "I just got a great deal on travel to LA".
Of the 7,000 tweets at LeWeb, how many are more than just one line assertions, opinions or comments? Given the nature of Twitter, you can't have much more than that.
You can't say carte blanch that Shaq is an authority based on his number of followers, but to the people who follow him, they can at some level consider him an authority. That is why they follow him. If the other people they follow also follow Shaq, then his authority is even bigger to that person.
I think it is a technically feasible problem, but very computationally intensive. You'd have to calculate a different authority value for each pair of follower relationships in Twitter, and that is many orders of magnitude greater than the number of Twitter accounts.
- Daiv http://Twitter.com/DaivRawks
This is usually measured by what percentage of your tweets are answers to other people. The more conversation you are the higher the probability that you are listening and reacting to what people are saying.
Tweetstat helps with that, so does twinfluence. There are a few sites that help evaluate someone's twitter influence. The question is how to integrate that into a search.
I unfollow people who send unnecessary @-messages, much preferring to hear people's original thoughts and not the so-called conversations. it is completely useless to send the following tweet:
@so-and-so what?
What use is that to your other followers? Twitter is not IRC, it's not a chat room, it's not IM. Have conversations, yes. But don't make that your litmus test for who's a "good tweeter."
Numbers do not equate to relevant and meaningful content... anyway you slice it Perfect example: Shaquille O'neal (http://twitter.com/THE_REAL_SHAQ) has over 22,000 followers on Twitter. Should he have authority over other folks? absolutely not. Should I listen to him when he speaks? Well, I'll let you answer that one.
I think overall an authority line would help only in some cases ( like the predicament you have above). But if they install that feature- people will be even more incessant about gaining more followers (Which I believe is not the point of Twitter).
Regards,
Taylor Barr
A really intelligent authority algo should included twitter ratio, number of re-tweets the user has generated etc to start with. It could further include median of authority of one's followers etc.
I get your example of number of followers influencing amount of social capital/ opinion, but that value falls away in terms of any kind of expertise on anything non-subjective.
An expert/opinion/heavyweight on Objective-C/iphone stuff has way less followers than any A-list social media pundit (who merely bought an iphone), but would be artificially penalized if following a strict follower=authority rule.
Just saying!
Of course if you just want a brand to recognize your name and offer you a special deal then numbers matter. Brands do like celebrities to use their products and mention their names.
What I'd like is to filter an arbitrary set of tweets from an arbitrary set of people by intersecting them with people I follow and my tag cloud. Then again people in hell want ice water !
http://www.techcrunch.com/2008/12/24/twittangle...
Get over yourself loic.
Here you could search at least who in town is an authority on twitter!
http://twitter.grader.com/index.php?Action=Twit...
I am only No 3 in Mannheim, Germany
Sorry, but I deeply disagree with this, Loïc. #of followers doesn't say anything about how a person may be revelant for "YOU"... or not. What gather us online is our common interests. I've shown to Robert something that might interest you, and will gladly show an alpha version of it to you too :)
If you view Twitter as a platform authority should not be solely determined from where the conversation takes place.
Some more thoughts on that http://bit.ly/VcuZ
(I realize that's not the point of the post, but it's an interesting factoid.)
You are not thte only one and it's a big matter with web 2.0
See my article on this subject here (in french) :
http://archivesic.ccsd.cnrs.fr/sic_00122603
And on slideshare :
http://www.slideshare.net/oledeuff/autorit-vers...
http://thenoisychannel.com/2008/12/27/loic-le-m...
As for Twitter...I agree with Scoble: Too much else to fix first.
this is a terrifying and horribly flawed idea. you need only visit adherents.com to see the travesty of such an idea - go take a visit and tell me, based on number of followers, which religion matters most...
you are manufacturing problems in search of solutions that are in want of purpose...
You always have to consider the 'who' behind the numbers as well. Shel Israel once used the example of someone only having three followers, but those three followers being three world leaders. I've used the example of a social network where the owner has only a handful of friends, but can shut the site down anytime they want. Etc., etc. But, even with these examples, there is no link between credibility and the number of followers.
All my best,
Rich
What he is saying is simple math. If someone has followers in the tens of thousands then more eyes are going to see it. I think some of you are seeing elitism where there isn't any.
"A 'I just love this blog, and this post is why' tweet from a Twitter user with 1,000 followers almost always trumps a tweet from someone with 10K followers, that puts no context to the link."
I'm not so sure that's true, but I will say that if "A-list" tweeter and "D-list" tweeter both tweet just a link then I would think the A-lister would generate more hits and if they both put in the same context then again A-lister wins. That's all the writer seems to be saying.
Nothing wrong with that, but it again calls out the two different sides of social media: personal vs professional benefits. Does an individual looking to make one-to-one connections care about how popular their new friend is? An organization strategically building a network for disseminating information does.
Twitter seems to be the darling of the social media set because it can work for both of these audiences, but the trade-off is that it's not perfect for either one.
The kicker? He was more 'relevant' when the Bush Administration was wielding actual power. Limbaugh still has the same base of followers (and is being paid commensurate with his audience) but the current mood in the country automatically devalues his opinions.
So authority, reputation, linking, *and context* matters. The first three can be measured...but what about that last one? How are we to know that some tweeter isn't all over last month's thought leadership? Or that someone has a lot of followers because he was with a big software company and co-wrote a book, but is a mile wide and half an inch deep technically? Or even that tweeter is trying to influence or direct some initiative?
As one of my mentor professors once said, "Always consider the source, their motivations, and the context within which they're pontificating." Wise words.
Authority would be nice BUT how about letting us make different folders for different types of followers?
"If someone has followers in the tens of thousands then more eyes are going to see it."
So if I want as many people as possible to know Sprint is out on Twitter and looking to help customers, helping someone with as many followers as Loic would be a good move. Keep in mind that the commenters on this post have a much greater understanding of Twitter and authority than most. Right or wrong (and I happen to agree that it's wrong), the majority of Twitter users are going to see those with the most followers as having the most authority.
That said, my PR colleagues and I have also reached out to a lot of folks in the past who have very few followers. Our selection of who to help isn't nearly as scientific as you might think. Ideally Sprint would have an entire team devoted to social media and a Chief Twitter Officer like Southwest does (I believe). Or we'd have a person or group of Care folks solely dedicated to helping on Twitter like Frank Eliason (@comcastcares). While Sprint has given me free reign to “try new things” when it comes to social media, we're not where Southwest and Comcast are yet -- hopefully we will be soon. So instead, we have several PR folks monitoring Twitter for customer issues in their spare time.
I bring this up just to point out that one of the reasons @dotben may not have gotten a response could have been that neither I nor my colleagues saw the tweet when we were searching for Sprint comments. It really could be just that simple, and this is a great example I can bring back to my bosses and say: "If we want to fully join the conversation and engage, we have to dedicate enough resources to catch Loic's tweets, @dotben's tweets and the tweets of those folks who may have very few followers."
After all, how do I know that someone who only has 10-20 followers on Twitter doesn't have 1,000 Facebook friends or a large family and a significant social network he/she influences? I don't and I would be foolish to assume I can outsmart WOM. I'll lose every time.
What I do know is that the majority of corporate PR hasn't fully embraced platforms like Twitter as worthy of one or more full-time positions. It’s my job, as well as that of my colleagues', to make sure we convince Sprint that anywhere people are talking about your brand is indeed worth that type of investment and posts like this one actually help us make that point.
Just one more thing...Happy to help @dotben (I'll be tweeting to him right after this post) and any other people who have questions/concerns about our brand. So if you all hear from someone who needs help with Sprint, please send them my way (@JGoldsborough). Helping people with Sprint issues via WOM referrals can also help us show our leadership the continued investment Sprint needs to make in Twitter.
Sent from my BlackBerry® wireless device
Justin, that sums it up brilliantly. You don't need to filter searches by some kind of authority measure--if a message has legs, you'll know from the avalanche of conversation that it triggers. What is probably more useful are better, more interactive summarization tools than Twitter's "trending topics" that help you monitor the collective conversation without having to read each individual post. Particularly if you're tracking all of the social media, and not just Twitter.
Loic, you of all the people should know that nothing on the web lasts forever and that includes your twitter followers. Be careful when you ask for search by authority. It might turn out that by the time twitter implements this (they would do it only to satisfy your ego) you would not be having more than a couple of hundred followers.
P.S. Twhirl is the only reason I follow you!
I'm OK with Scoble's approach but I think the search by "authority" will deliver the same problems we have now with blogging - the best posts about a topic are not generally surfaced by authority measures. Instead, we get the most algorithmically appealing posts which are usually either a product of old A list bloggers sticking together and linking very opportunistically or overly SEO'd posts that suck but do a great job fooling the algos. Mostly ranking is now a combination of those two factors (old stuff and SEO measures).
One of the *great* things about Twitter is that it limits exposure fairly democratically. Authority search will help the twitter "rich" get richer, but I hardly think that's a noble objective - it's the same problem we have now where early adopters with a superficial voice are elevated above quality journalists.
Unless I'm missing something it sounds like you and Mike want to make sure Twitter does not threaten the status quo with more democratic ranking. I think it's a great idea. In fact I think it would be interesting to *reverse* the algo you suggest - I'd rather hear from some Grandmas in Peoria about their iPhone experiences than from Jason Calacanis about [groan] the wonders of Mahalo.
Unfortunately, the US is a country of soundbites and you've got to learn to choose your first 10 words more carefully because many people won't read beyond. The entire reason search engines deal with algorithms is because we need ways to search through vast amounts of data to find the most relevant information.
It would appear that most of the people commenting here and on TC are ignoring the fact that Twitter is rapidly reaching a point where their message volume requires sorting through additional ways than just reverse chronological.
Fight the good fight.
If I read this right, now you want to be able to weed out the noise...? What if you are the noise?
Just sayin'
http://twitter.com/pixelbug
In summary, your suggestion would militate against subject matter authority, by skewing the network toward greater network centrality in an essentially content-free way. I would encourage you to think about how this kind of feature can be counter-balanced so those who are less-linked can have their voices heard. This is especially needed in cases where authority matters.
Really, people ought to focus less on connectedness of central figures than on connectedness of those at the periphery. Unless you think the most-linked and most-read really are expert in just about everything, and others' voices shouldn't be heard.
The "twitterverse" needs to get OVER itself
The experiences and perspectives of those in toprank positions tend to be very different from most people. Check out Malcolm Gladwell's latest work for some explanations as to why.
Yeh, I too am against having an authority search.
No wonder we get lost along the way, and you've got a hell of a job in finding out who really does know something about the issue you're pondering. I don't have Loics reputation, so I don't count as an authority here, but I dare say Loic is wrong. He's off by miles.
In his blogpost he pleas to have a twitter search by authority, just like technorati who worked out an algorithm to define the authority of your blog. I'll walk a mile with him on this path as the algorithm to define authority by number of citations or links is much better than counting sheer numbers of followers. However...
There's a catch.
The catch is in creating an elite layer, the twitterati, the digerati, or whatever you'd like to call them. While reading Loic's blogpost, two things came to mind. First a conversation I had past New Years'Eve and a blogpost I wrote about half a year ago on the Social Web.
The New Years Eve conversation I had was a conversation with my neighbour and my Sister in Law who has recently received her PhD at the Oxford University. She graduated in the interaction between insects and how that would affect a Ecological systems or something like that. Fact is she worked at the same department as Dawekins (the Evolution theory zealot) and the discussion went into Evolutionism vs. Creationism. On both sides you have zealots and with neither you can have a normal scientific, fact based discussion. Evolutionism is the dominant philosophy in Science these days and to most people it seems like the case has been closed. Evolution has been scientifically verified, beyond doubt. Well, it isn't. I didn't see a video on YouTube to prove it (nor did I see a video on YouTube to prove Divine Creation) and if you would conduct objective, unbiased science, you would have to conclude that the evolution theory has gaps. In a scientific setting you'd count on educated minds questioning these results, but in the way it is presented to our children who do not have the cognitive skills yet to analyse results, we are brainwashing them. If you look at how the scientific scene works it explains a lot. Authority in Science comes from the number of publications you have in a major magazine. Every paper you submit is reviewed by an editor who likes it, or not, regardless of the argumentation to your findings. Let's say you write an article about how Evolution sucks, no matter if you include 100% proof, if the editor doesn't like it, you're out. Next step is the peer review. Every paper, once it has passed the editorial selection, is sent to peers, colleagues and the same selections starts over again. Let's say my findings are solid and proves the previously published research of one of my reviewers wrong, he won't like that as it will make him lose his reputation, authority or stature. Case closed. No publication.
Selection and authority in this process kill Science as it should be unbiased and objective. It isn't. I think the same would count for authority based filtering. The key issue here is in automation. Google and other search engines have worked out algorithms, as well as technorati who put auhority to blogs. No matter how much intelligence you put into these intermediates, they cannot compete with the selection capacity of the human brain. These selection mechanisms will undoubtedly produce a prevailing elite, just like in the science case above and smart, intelligent and argumented opinions to the contrary will be neglected.
This made me recall a post I made several months ago on the social web called "Power to the Community" In this blogpost I discussed how my colleague defines social webdesgin. This is way more than defining the social web. It is about desinging your websites to create emergent behaviour. In extremis this could lead to Isaac Asimov's foundation series in which he presents the Psychohistory.
The basis of psychohistory is the idea that, while the actions of a particular individual could not be foreseen, the laws of statistics could be applied to large groups of people and used to predict the general flow of future events. Asimov used the analogy of a gas: in a gas, the motion of a single molecule is very difficult to predict, but the mass action of the gas can be predicted to a high level of accuracy - known in physics as the Kinetic Theory. Asimov applied this concept to the population of the fictional Galactic Empire, which numbered in a quintillion. The character responsible for the science's creation, Hari Seldon, established two postulates:
That the population whose behaviour was modeled should be sufficiently large
They should remain in ignorance of the results of the application of psychohistorical analyses.
In creating automated intelligent interfaces to filter through the inprocessable amount of digital information we might just be on our way to do that...
http://www.denisflorent.fr/twitter-authority-al...
# of followers if the only metric that most Twitter directories use, which makes it hard to find people that actually have something to say.
Some with 1,000 followers that only follows 50 people, interacts with lots of people, and loads of people RT what they tweet will currently rank in Twitter directories below someone with 2,000 followers that followed 3,000 people just to get to 2,000 followers that doesn't tweet anything interesting or participate in the community.
Does that make a Twitter directory based on # of followers useful?
I would say not.
Would someone please build a quality-based Twitter Directory (even just multiplying # of followers by a quality score would be a huge improvement (# of followers / # following).