Curiouser and Curiouser!

Curiouser and Curiouser! on leaderboardism http://matt.blogs.it/ RSS feed for topic leaderboardism Copyright 2006 Matt Mower Squib/0.1 self@mattmower.com self@mattmower.com en-gb Weave a circle round him thrice Mon, 22 Aug 2005 14:12:22 +0100 I just <a href="http://www.feedblog.org/2005/08/kottke_is_upset.html">Kevin Burton's response</a> to <a href="http://www.kottke.org/05/08/so-long-technorati">Jason Kottke abandoning Technorati</a> in which he says:<blockquote>I'd rather have a Technorati that was fast and always worked even if that meant only indexing 1M blogs. Even 500k blogs as long as they are the top 500k blogs.</blockquote>Which is, I think, indicative of a class of problems people are experiencing in thinking about the blogosphere that revolve around a concept I'll call Leaderboardism. Right now Technorati are <a href="http://www.technorati.com/about/">claiming to index 15.7 million blogs and have a database of 1.4 billion links</a>. WOW! Those numbers are certainly impressive. But what does this huge data-warehouse buy us? Gripes about performance and database outtages aside, not much it would seem. I don't get anything from a Technorati search that I value over, say, a Google search. In particular I don't seem to get value from Technorati understanding the blogosphere better than Google which you would think they really should. Kevin thinks a better idea is to just index the most important 500,000 (3% of Technorati's claimed reach) of blogs in the blogosphere. Sure that would make Technorati fast. But would it make it more useful? After all, who is deciding who is important? How are they deciding it? And isn't importance subjective anyway? To my way of thinking what Kevin is advocating would make Technorati faster and less useful in equal measure (unless you are mainly interested in what the usual suspects think). The problem is that the blogosphere has grown too large for summary statistics to be relevant to a large group of people anymore. Your Top 100 isn't mine because you aren't interested in basketweaving and vole racing and I am. I think Technorati (and <a href="http://www.feedster.com/">Feedster</a> who seem, so far, to have avoided many of Technorati's pitfalls) should abandon Leaderboardism and focus instead on how to make their database relevant to each individual. Relevance is about understanding the context of the reader and delivering the results they would have asked for if they'd only known what they were. I will consider it a success not when I can see <a href="http://www.technorati.com/pop/blogs/">The Top 100 Blogs</a> but when I can see The Top 100 Blogs you've never come across but will wish you had! For reference I read 2 of Technorati's Top 100 (although I have read about 30% at one time or other and am familiar with over half) so clearly their measure of relevance doesn't match mine very closely. This takes me all the way back to where I started thinking about <a href="http://matt.blogs.it/2002/05/31.html#a63">Village Shops in Blogspace</a>. Update: This post lead to me being asked to contribute to an <a href="http://www.publish.com/article2/0,1895,1856550,00.asp">article on Technorati</a>, and later to my <a href="http://matt.blogs.it/2005/09/08.html#a1981">posting a follow-up</a> item. http://matt.blogs.it/entries/00002867.html If I may presume Thu, 08 Sep 2005 17:03:50 +0100 I was asked to contribute to a Publish.Com article, <a href="http://www.publish.com/article2/0,1895,1856225,00.asp">Five Reasons Technorati Is Broken (And How To Fix It)</a>, by Jason Boog that went live today. Inevitably in such situations you write a lot of stuff which gets edited down to a few paragraphs that, when you're lucky, represent some aspect or other of your thinking. In this case the brokenness of Technorati isn't that interesting to me because it's not a service I use. What is interesting is the opportunities that may be before Technorati to deliver an altogether different kind of service. Whether those are real opportunities or not is hard to say right now, but I feel it's worth exploring the topic. In the article Jason quotes me as being against popularity contests (which I am) and, instead, interested in finding the blogs nobody knows about (which, strictly speaking, I am not). As Terry Frazier and I have <a href="http://www.terryfrazier.com/1757">discussed</a> on several <a href="http://www.terryfrazier.com/1748">occasions</a>: it's not being unknown that counts, it's being relevant. What I want to know about and track are blogs I don't about which are, nevertheless, relevant to me. I just don't think Technorati are doing that. Could they? Taking it to the next level <a href="http://www.drmartinhall.com/">Martin</a> and I are looking at what lies underneath relevance. We think it's values. So we're working on bringing value to the process too. Anyway, I spent some little time thinking about what I wanted to say and, with Jasons permission, am posting here both his questions to me, and my answers: <blockquote> <dl> <dt>1- Could you define "Leaderboardism" and how it relates to Technorati? It's my favorite term to come out of this whole Technorati debate...</dt> <dd>Leaderboardism is, in my opinion, a misguided obsession with blog popularity rankings. Knowing about links into my posts is useful, it helps me understand what is going on around things I talk about, but it doesn't scale from there to the blogosphere. We all know by now that 'boing boing' is the most linked weblog in the world, so what? I don't read most of the "top 100" because they're not interesting to me. I want a heads up on new blogs which are interesting to me, i.e. the other end of the spectrum from the 'boing boing's of this world. Why do I only come across them by chance? And how can we change that? Tagging is a part of the answer but naive solutions like Technorati Tags just end up creating an even bigger mess. My view, since about May 2002 when I wrote a post "Village shops in BlogSpace" (http://matt.blogs.it/2002/05/31.html#a63), is that we need to get at the meanings, for individuals, of what's going on in the world. Make the Blogosphere revolve around the individual not the other way around. My original intention had been to create a service which would "understand" the meanings underlying what people wrote and to use this as a way of match-making bloggers who didn't know each other yet but who *should* know each other. Through this process I hoped to form (and reform) fluid communities of bloggers with common interests. However the technology wasn't up to the challenge. So instead I started to look at ways in which bloggers themselves could tell us about what they were interested in. I wrote a tool, liveTopics, and started adding topics to my blog posts around June or July of 2002. Unfortunately I hadn't even begun to realise how much I had underestimated the challenges involved. </dd> <dt>2- If Technorati hired you tomorrow to "make their database relevant to each individual," what would you do?</dt> <dd>I'd start with a recipe along the lines of:<ol><li> <li>We want relevance to the individual => we need to understand what is relevant to the indivudal, so</li> <li>Allow individuals with an account to register their interests (either directly or indirectly through their writing)</li> <li>Develop relevance metrics for blogs with respect to tags</li> <li>Cluster both sets</li> <li>Experiment with relating clusters from each set together</li> </ol> This is kind of mixing my vision of a service with what I know about Technorati today and may not work really well. In my original vision strong clusters could be named you could then begin to offer services to the people attracted to them. </dd> <dt>3- Terry Fraizer's blog mentioned that you helped develop some tagging and blogging tools. Could you give me a brief rundown of that work experience, and tell me how many years you've been working on the problem of organizing text sensibly on the Internet?</dt> <dd>It started as I was deploying Livelink, a big knowledge management package. It was powerful software but, at the same time, something was missing. Then I spent some time working in software development and came across the eXtreme Programming ethodology. When I startedblogging and saw what was possible when anyone can write for themselves it just felt right. It occurred to me that traditional KM had become detached from the needs of individuals and I evolved a kind of "eXtreme Knowledge Management" philosophy to combat that. From that point, mid-2002, onwards I've been building tools to try and address this. I wrote liveTopics for the Radio Userland blogging platform. This was used by quite a few people (e.g. Lilia Effimova http://blog.mathemagenic.com/) but was ahead of it's time and, frankly, on the wrong blogging platform. Radio was dying (though I didn't realise it at the time). I went on from there to co-author K-Collector at Evectors (http://www.evectors.com/). K-Collector was much close to my vision. It was a combined blogging platform and aggregator we released in 2003 which allowed a group (e.g. a team) to collaborate using blog posts, topics, and relationships. We did quite a bit of work on developing a simple topic model which gave more ability to express meaningful relationships. I left Evectors in July'04 and now work for a company in the privacy and identity managament space. However I also study part-time for a postgraduate diploma in psychology and am very interested in Social Psychology and the Psychology of Personality. As part of my study I am working with Dr. Martin Hall (http://www.drmartinhall.com/) on a project to understand more about how people express their value systems through blogging. This is where I think the next break through will occur. People need information but, underneath it, they want meaning. The joy of weblogs is making that connection with other people. When you first shake hands with someone whose blog you've been reading you aren't meeting for the first time. You've already had a meeting of minds and know a lot about what makes each other tick. You share interests but you also share values. We gravitate towards the things we think are "good". Martin and I are working on bringing more of that to the blogosphere. </dd> <dt>4-Is there anything else you want to add about problems with Technorati?</dt> <dd>I think I've already said publically what I think is 'wrong' with Technorati from my perspective so I would just like to add one caveat. I don't know David Sifry but he seems like a smart guy and he has smart people like Kevin Marks, Doc Searls, and David Weinberger around him. I don't see what I believe to be important reflected in the Technorati business model but, for all that, David is the guy actually out there running a service.</dd> </dl></blockquote> http://matt.blogs.it/entries/00002889.html