Filters

Call My Agent: Evolution In Information Retrieval
Call My Agent: Evolution In Information Retrieval 772 457 Raymond Blijd
 It is known by many names: OffspringMagnets, Filters, Bloodhounds  but in theory they can also be called: Agents. My definition: An intelligent application that basically goes and fetches ‘your’ information without you re-entering a query at every instance. Undoubtedly there are other more scientific explanations but for argument’s sake we’ll keep it simple and stick to this one. Now, how can agents make our lives more pleasant?

 

You’ve got Mail!
The first ‘agent’ I used online were email newsletters. If an interesting site had the option to subscribe to an email newsletter it gave me two features in exchange for my email address and exposed preference:

  1. I did not need to visit the site to get the latest information;
  2. By popping up in my inbox, it was a reminder that this source existed.

More advanced email delivery systems gave you more flexibility in delivery times or topics to choose from and the more advanced sites gave you more sources. Yet, not all sites had these options so I moved on.

webtimeline

Blowing the Pipe
With the rise of R.S.S. (Really Simple Syndication) feeds also came more creative features to play with. As with email one could subscribe to a site and use a separate application (RSS Reader) to receive the information. Yahoo Pipes – An interactive feed aggregator and manipulator- gave me the extraordinary ability to, “rewire the web”. For example, one of my favorite pipes was created right after the launch of the first iPhone. It used several online feed sources and persistent searches to scour information on the iPhone. If a site did not provide a feed I could always create one for it by using a web scraper.

While it all started out nicely, information overload crept in really fast. By using services like Postrank to prefilter my feeds based on popularity I was trying to stem the tide. By using trends and suggestions instead of Pipes in Google Reader I was also encouraging serendipity.

Back to Basics
Twitter, Friendfeed ,and Facebook added a new dimension to this overload fire I was fueling. Especially FriendFeed was menacing in its torrent of good information. By piling on filters on feeds I think I’ve managed the problem. My latest darling filter isSummify iPhone app. It summarizes and reduces my Google Reader 78 feeds and Twitter 226 friends to 10 story-links every 8 hours, depending on my settings.

Looking at this from another angle I think I have subconsciously created a personalized newspaper edited by humans and machines alike. Yet, I feel like I’m still missing some things in terms of relevancy, quality, and serendipity. Moreover, in the real world, I’m still pleasantly surprised by learning new stuff from friends and colleagues at the water cooler. It’s also scary to realize that speed will increasingly become more important. The earlier we know something the better prepared we can be and you might even be able to predict the future.

I think the web has evolved enough for us start rekindling this notion of agents. My dream is to have an intelligent agent which can rummage the web and alerts my friends in the real world to provide me with relevant information based on my profile. In real time or at predetermined times based on the substance of the information and the impact on my live and work.

More importantly, I hope that the web and my friends also know that I’m available to share, just ring my agent.

Burning Filters and Popping Bubbles: The Personalization Paradox
Burning Filters and Popping Bubbles: The Personalization Paradox 834 472 Raymond Blijd
 In 2008 Clay Shirky coined the phrase “It’s not information overload. It filters failure.” In 2011 this was reversed by Nicolas Carr stating: “It’s not information overload. It filters success.” He also stated that we were, in fact, getting dumbed down by filters like Google. Eli Pariser at TED gave a presentation called Beware online “filter bubbles” which gave a warning of the dangers of filters and personalization. So are we going filter mad?

bubblesA question of perception

The reason for the filter upheaval is understandable from a certain perspective. So was book burning in its day. Filters have been around since there was information exchange of any sort. Eli Pariser made the case that in the last century we had filters called editors with build in ethics who decided to give us a wide spectrum of information. Algorithms – without ethics – are replacing editors in this century as our main filter. However, the problem is that algorithms for personalization will inevitably calculate your persona and provide you the information it can deduce you might want. This can be very confrontational. But that’s not much different from the previous century whereby the act of buying and reading a certain newspaper brand is highly ‘personal’ and would say a lot about who you are.

bubbles2The luxury of serendipity

In a world of consumer-targeted filters, do these also pose a “danger” for the business market? On the contrary, they are the core of the business proposition. The rise of information providers and publishers depended solely on finding the appropriate filters called sources, experts, authors, editors, etc . A lawyer arguing before the bench, a physician at the bedside, or an accountant before a filing deadline do not have the luxury of the accidental discovery of crucial information. They rely heavily on meticulously chosen filters to deliver in a clutch. And now they are requiring us to design better filters and not force them into haystacks to find a needle.

Information overload is a symptom best cured by filters that do not aggravate the problem or deceive the user.

bubbles3

Content is King, Search is Queen and Filters Are Their Offspring
Content is King, Search is Queen and Filters Are Their Offspring 1024 557 Raymond Blijd

In a digital space, all content providers fight for user face time which makes it increasingly difficult to be exceptional. Although good branding and premium authors help you stand out on a bookshelf, additional features on top of content will tip the scale in a digital environment. Content providers should, therefore, move towards providing solutions instead of just information. But how?

Search vs. filters
If search is your game plan the approaches might vary from finding the best authors on the subject matter and manually or semantically tagging their content before the search, or understanding the user behavior and adapting the content after the fact (see Google’s Marissa Meyers on “contextual discovery” – starts at 4:23).

I must admit, search is usually the first thing most people turn to, but only due to a lack of better ‘filters’ available. So what might be a better filter than search?

One obvious filter is to ask a person – usually a peer or colleague. Filters currently serving this purpose on a grand scale are Twitter, LinkedIn, and Facebook. Facebook overtook Google as the most visited site in the US for 2010. Also see Forrester’s research, Peer Influence Analysis.

A less straightforward filter is www.citebite.com or www.snip.ly: a tool for bookmarking any passage inside a web page and sharing the direct link to the passage with others. In legal research, citations originated from paper perspectives and the rules, therefore, are bound by it. Yet these bookmarking services prove how a digital platform can greatly enhance an archaic system of citing materials. Imagine a future where such a service would be the standard for digital references and citations.

 

legalcomplex filtersA content agnostic view
There are many filters conceivable to assist a knowledge professional in finding what they’re looking for. To find what these tools starts with having a content agnostic view. Taking a content agnostic view means looking at the fundamental tasks knowledge professionals apply in their daily work. What actions do they take in order to ‘know’ things based on the information accessible to them? Moreover, how can we turn those tools into online services? See more examples of filters as online tools in this mindmap (click it to enlarge).

 

Knowledge_Library legalcomplex2William Gibson once stated, “the future is already here.” Most tools already exist. Yet they aren’t being considered for a professional setting because of their often mundane consumer market origins. What we are looking for is most often hiding in plain sight.

Tags for this article: