Attention!
Information overload has been and will continue to be a growing problem due to the natural growth of the web and the exponential growth of valuable user generated content such as blogs. New technologies and methodologies to deal with it and I believe we’ll see a wave of solutions hit the web in short but due time to deal with it.
My friend and colleague, Daniela, was the first to bring to my attention (no pun intended) with her post on Touchstone, now called Particls, which is an application leveraging APML. APML stands for ‘attention profiling mark-up language’ which allows an application to build a dynamic attention profile based on what you look at and read.
The idea is interesting and Particls is making some headway. I am an alpha tester of the software and while intrigued, I’m still not sold on the idea as of yet simply because I think for it to work, you have to have an engine tuned quite well for it to predictive of what you want to look at based on what you’ve done historically. I believe the ultimate question is do most people really read the same topics or themes all the time?
Chris Saad, founder of Particls, had an interesting post yesterday on attention management and Web 3.0. While I think attention management software may find its place in the overall web scheme, I agree with Chris’ point that it is not Web 3.0.
First I agree with Chris’ point attention management software has its place in the current web, it processes what you read and puts up-front-and-center predictively what you would want to read. Secondly, I believe the Web 3.0 is going to be the ‘Semantic Web’ or the semantic web will be a very big piece of it. We simply need to get to the next level where the data within web content can be utilized at the data level and not the document level. Let’s take the leap where APML extends to profile your attention on the data within documents and not just the documents, at that point APML would be very Web 3.0.
To add to the confusion, and to give extra technical headaches for anyone trying to find a filter for them, many people don’t read online anymore, they seek video or audio podcasts to explain it.
Lou - thanks for the mention
I agree that we will live and die by both the accuracy of the Personal Relevancy algorithm and the simplicity of the training.
We are only about 50% of the way through that process for version 1 - And, of course, it will be a constantly evolving process. Chris Messina is in the workgroup and will surely be helping with that.
You are right that APML would be part of the semantic web. We are also interested in combining it with Microformats so that it can be embedded into pages etc.
Thanks!
Dirk, I agree that video/audio is being a burgeoning meme on the web and it will present a challenge to both filtering and attention technologies. But I think we are still some time away from where people turn to video/audio as their predominant mode of consumption, even with the huge penetration of the ipod. These technologies, nonetheless, will have to adjust.
Chris, thanks for the follow-up. I look forward to seeing the progress that you and team make as you continue your roll-out and improvements.
BTW, is there a way to increase the amount of time that my scroll keeps the headlines active. Often I am working and don’t have time to pay attention and I worry that time scale causes the system to “think” that I’m not interested or “attentive” even though I simply didn’t have time to click.
@Dirk.
Video and Audio are valid forms of media consumption that require Attention management. While I agree with Lou that for busy people text is far more efficient than video (you can skim text quietly while you have to watch video presentations their pace).
The neat thing about Particls is that we leverage a lot of existing technologies to bring the pieces together so we can (and already do) leverage a lot of existing video/audio transcription services and metadata to rank and discover audio/video content as well as text. Probably more so in the future.
@Lou
There is no way to change the time an item stays on the ticker (yet) as time is actually not the only factor that removes it. We use our Personal Relevancy algorithm to determine when an Item should appear on the ticker, and we also use it to work out when the item should be removed.
That being said, however, items that are not clicked on do NOT count against your interest in them. Clicking on items counts your interest FOR them however.
Not having time to click is perfectly ok by the way. The ticker is designed for just such a scenario. You are still seeing headlines without clicking - that’s it’s power. And if you ever find it distracting or if something very important happens, there are always the popup alerts.