Get a little anxious, information overload, tagging

This is a braindump of something I would like to think about and possibly approach some solutions this year.

Information overload makes me more than a little anxious. I find it amusing and also sad that a search for “information overload” on google yields over 11.2 million results.

The problem is that there is so much good stuff out there but no really useful tools to help sift through this meme avalanche. How does one sift out information from noise, and more importantly, how does one turn that information into usable knowledge?

After all, I don’t really care about reading 10,000 blogs, but I sure would like to pull out some useful tidbits that will improve my life at the moment I need them.

(Fair disclosure: I’m doing some consulting work for Pluck, the company behind the social bookmarking site, http://www.shadows.com and the publishers of the Pluck RSS reader. A relevant competitor is the Yahoo company, http://del.icio.us).

One useful tool that has been all the rage among the hipsters is tagging (tagging has been around FOREVER, but reached a tipping point in the last couple years when services like flickr.com showed their usefulness).

Tagging lets you associate words (i.e. “tags”) with nuggets of information. If I am reading a website about moleskine artwork, I might decide that the site should be “tagged” with keywords like “moleskine”, “art”, “journal” etc.

Services like shadows.com allow you to add these tags to your bookmark collection. So later, when I want to do research on notetaking, I might do a search among my bookmarks for those items tagged with “journal” as a keyword.

My search would result in all those bookmarks tagged with “journal” including my moleskine art site.

In a way, you can think about tagging as putting your information (in this case, bookmarks) into file folders with arbitrary names. The difference is that you may assign that same information to lots of different folders.

This works well because you can now put your information into multiple categories very quickly. You can also run statistics over your tags to determine things like “what are my greatest interests (i.e. what are the tags with the most associated items)” and “what tags are closely associated with each other?”

The flaw is this system relies on human categorization of information. This is a manual process that is potentially time consuming: find, read, tag, find, read, tag, ad naseum. Computer artificial intelligence is not nearly close enough to being able to automatically tag information reliably.

Now there are companies, like Pluck, that have developed solutions for helping spread this work out among communities. They expose your collection of tags to everyone else. So when I do my search for “moleskine” not only do my bookmarks show up, but optionally, I can also see the bookmarks everyone else tagged with “moleskine.”

These are great steps. They help humanize the web, but have also turned us into a society of robot librarians.

So the important question is, “to what end?” I don’t think there has been a lot of detailed work in synthesizing this information so that it is usable knowledge.

In a typical web surfing session, I will use my rss reader to find interesting websites mentioned in articles on peoples’ blogs. Often, if this site is interesting, I’ll tag it through shadows.com with some keywords and then move on. Because there are so many more sites to explore, I don’t spend a lot of time reading. It’s more like, “skim, identify enough information to generate useful tags, tag, move on.”

My assumption is that I will be able to go back and search through my collection when I need the information. The reality is that I’ve spent an hour or two with low-thinking, and ultimately, low-impact work of tagging.

This is not a valuable use of my time.

Worse, this tagging will never conclude. For all intents and purposes, as it pertains to human capacity for learning and remembering, the internet is infinite. I will never run out of links to explore. When am I “done”? Chasing links is a never ending pursuit. We are stressed by tasks that will never end.

An approach I would find interesting:

Create an organized research task-list or “call sheet” that helps focus and direct my reading and research.

First, I would keep an ongoing list of research topics. This might be a collection of tags that I would want understand better. I would schedule time to deeply explore, note-take, and think about these topics.

I would use my collection of tagged bookmarks as a starting point.

This is a potentially slow process, but would yield some deeper exploration of topics. However, the problem of an unending task remains; how will I know when I’ve learned enough about the topic? It compounds because as I continue learning, newer subtopics will reveal themselves. Futhermore, older knowledge I have obtained will distort based on my new knowledge, and might require a reexamination.

However, I believe we can develop heuristics and tools that will allow me to better prioritize what I should be reading and should be learning.

Questions I have

• How do you quantify what you know and how do you list what you don’t know?
• Can we build tools that automatically tag incoming information with reasonable accuracy?
• can we build toold that prioritize incoming knowledge collection?
• How do we “heal” our collection of knowledge so that badly tagged information is better appropriated?
• How do we “evolve” our collection by exploiting interrelated information?
• How do use metadata about our reading and searching habits on our own collection to highlight and prioritize certain topics?
• Can we create alternate representations of knowledge that are visually and/or kinesthetically based? See – ambient information devices
• Can we create a “radar” of information topics that function with more information rich data than tag clouds?