Relevance + Trust = Attention

I just read a few interesting posts by Tim Bray and Alex Payne about what to read and how to stay up to date (see below).  Much of what they say I agree with.  The simple problem is that there is just too much stuff out there that is interesting or important on some level.  Combine that with an ever expanding workload, a short attention span, and a fading memory and you have a combination that just can’t work long term.  What’s interesting is that I’ve asked several knowledge workers of one sort or another what their biggest problems are and most respond with something like …

  • “too many interruptions”
  • “wasting time on nonproductive tasks like email”
  • “no ability to focus on key tasks”
  • “excessive multitasking”

This is clearly a major problem and is probably getting worse given the increasing amount of information that keeps pouring in.  The key to solving this would appear to be to focus our attention on only the most important things.   But how is this even possible?

Attention is the most valuable commodity you control (at least from my perspective) and since I’m not one of those crazy-smart teenagers that can do 10 things simultaneously with full attention to each (my daughter apparently is though:), I need to manage my attention carefully.  The way I see it is what deserves my attention can be defined by a simple formula, something like Relevance To Me + My Trust In Source = My Attention (R+T=A).  The problem is that both Relevance and Trust are hard problems to solve.

Relevance is how pertinent, connected, or applicable something is to a given subject matter, in this case “me”.  There is so much available information today that relying on traditional methods of information discovery (search engines, RSS, etc) is just not enough anymore.  Many estimates put the number of knowledge workers in the US at between 28 and 45 percent and knowledge workers by definition require information to be effective.  Finding the right information to help knowledge workers excel is what Google’s about, right?  Well not really.  Search engines are great for finding the most popular items related to a subject but nowhere in the R+T=A equation does it say anything about popularity, so that’s not going to work.  Relevance is not a popularity contest it is very personal to you and should be handled dynamically based on many factors like whether you are at work, on a business trip, vacation, at home, working on an email, etc.  The people and places most relevant to you will be based on what you are currently doing.  Much like a spam filter learns over time what is good and bad so too should your relevancy filter learn what is important or not.  Is that Tweet relevant to the report you are working on or is it a link to a photo album of someone’s vacation?  If you’re having dinner in a restaurant should your Blackberry notify you of an incoming email?  Technology today allows many of these situations to be probabilistically identified it’s just a matter of breaking down some barriers and then SMOP (small matter of programming).

So that leaves us with Trust.  What makes a source trustworthy?  I think trustworthiness implies an existing relationship of some sort with explicit relationships having the most profound effect.  This won’t always work but may help initially filter the lion’s share of incoming data.  It’s interesting to think that a person’s social graph could be applied as a trust mechanism.  How well you trust a source depends to some extent on how far away from your node they are.  Hmmmm.  It would also be interesting to have contextual information associated with your social graph.  Given the dynamic nature of relevance, if social graph information also contained contextual information about the commonality shared by two nodes this could be very useful.  For instance if my social graph contained information about “how” I was linked to someone via work, local community, interest groups, etc, this could perhaps be used to help filter results given my current situation (ie. on vacation).  The relevance of certain connections could be elevated given my situation thus, changing the importance of the information.

In the link below Alex argues that his social network is incapable of helping him find the most interesting links.  I agree with that to some extent.   I certainly don’t think that what my social network reads is all I should read.  What would be the fun of that :-) ?  However I do think that my social network can be a part of the filtering mechanism and help with trending topics or group related research.  Again, I think it goes back to the dynamics of relevance.  If the topic is relevant to what I’m doing and my social network is finding useful information about that topic then maybe its worth checking out.

Trust without an explicit relationship is much harder.  Once a node in the social graph is not specifically connected to another node you would be relying on someone else’s trust in a node which is tenuous at best.  However, again I would say that even a degree of trust is better than none.  If the person I have an explicit relationship with trusts this person then maybe I should too (a little)?  Obviously this doesn’t work as you get further away or you’d trust everyone, but maybe there is some measure of trust in these extended relationships that can help us discover relevant material.

I’m not sure where this will all lead over time but I do know that the current state of information overload is unsustainable and will eventually be fixed by someone, maybe you:-)

Facebooktwittergoogle_plusredditpinterestlinkedinmailFacebooktwittergoogle_plusredditpinterestlinkedinmailby feather