Autonomy Just Doesn't Get It

Below is an interesting snippet from ZDNet. It's difficult to know where to start -- is the Autonomy position that del.icio.us and Flickr don't work? Perhaps someone should send Mr. Lynch a copy of the book Ambient Findability or setup a conversation with Thomas Vander Wal on the value of tagging. I do not find clients saying that they are going to abandon search and taxonomy tools. Most customers and other industry experts I talk to are looking at tagging, social bookmarks, tag clouds and folksonomies as solutions that compliment and add value to their information architecture efforts. It is obvious that people will tag for a variety of reasons (both good and bad) and also does not take much insight to note the problems associated with tagging. And yet consumers and companies like IBM (and others I cannot mention due to confidentiality reasons), have found significant value in these tools. IBM has been involved in several webinars in the past discussing the feedback loop folksonomies create that benefits those involved in corporate taxonomy efforts.

This does not diminish the need for tools provided by vendors such as Autonomy - the solutions are synergistic and not exclusionary. Andrew McAfee, Associate Professor, Harvard Business School has also documented an example of where a tagging solution can improve findability. What Autonomy should do is not complain about the evils of tagging but actually deliver a tagging companion product that is integrated with its traditional search and taxonomy tools. While up on stage, the message could be much more persuasive and influential - positioning Autonomy as a vendor working to advance the broader issues around "findability". That would show some market innovation rather than giving the appearance of protecting a status-quo business model that perhaps is not keeping up with the times. 

During a presentation, Mr Lynch slammed the popular practice of tagging web content and says that it won't help to organize information. Mr Lynch quoted an essay by Cory Doctorow, the science fiction writer, titled Metacrap. "Tags don't work because people lie, they are lazy, and they use different tags. And there is a huge amount of information that will never be tagged."

Source: » Autonomy CEO says tags don’t work | Tom Foremski: IMHO | ZDNet.com

Comments

[Scott Quick]

Mike,

Perhaps you just don't get it and history will judge tags a complete waste of time. Every word, symbol and picture that is used in human communication is a tag. Used together and taken in context, they create meaning. When search engines are perfected in the not so distant future, tagging will be looked upon like some bizarre voodoo-like ritual. Google and Automony have it right.

As for those companies engaged in tagging (including Attensa, in which I am an early round investor) they are simply following a fad which will ultimately fade. The business value these companies claim to be providing will ultimately fade as well.

Example... what was the name of that file and where did I save it? Forgetting something as important as a filename and where it is saved is the ultimate example of why tagging (ala naming) something fails miserably. Perfecting search engines is the only cure for our imperfect mode of human communications (and memory).

- Scott

[Mike Gotta]

Most search engine vendors for believe that their entire value proposition is in the content - they completely neglect the collaborative and social aspects that happen around the information - when I attended a search conference down in NYC last May, the only vendors that saw the needed balance between "indexing" and user-generated tags were IBM, Microsoft and Google. All the other vendors simply had a blank stare.

If I want to "paint" an object with certain tags so that it has meaning to me - how are search engines going to crawl my head?

Sure, search engines will better at indexing the context and improve on entity extraction, clustering, and so forth - but they cannot predict how I want to assign "meaning" to the information in a manner that only I know at the time I want to tag it. Take a look at where Microsoft is heading with its search efforts that accounts for social distance and what they are doing with Knowledge Network - look at what BEA is doing with ActivityRank - look at what IBM is doing with Connections.

Your example is faulty as well. A search engine should be able to not only locate a file based on its embedded content and any internal meta data (any desktop search will suffice) but on its assigned tags - if you look at some of the demos FAST has done - tag search results (i.e., de.licio.us) appear alongside traditional search listings - so this is about a blended, multi-facted search approach - Siderean is also thinking along these lines as well.

[Scott Quick]

The meaning you ascribe to content changes from moment to moment. It does so because of influences of time, increased awareness, changing context, etc...

As for painting content with tags ... why use metadata (an abstraction of the original content) to re-find it again sometime in the future. You found it in the first place, so what, you won't be able to find it again? Not a compelling argument.

And, are you telling me that a computer system cannot sit above a crowd, watching and analyzing for patterns without tags? Better not tell that to the boys at DARPA or NSA.

As for your examples of MSFT, IBM, et al... they follow any trend that has commercial viability - wheather or not it makes good sense in the long run is another matter entirely.

[Mike Gotta]

I am not saying search is not important and I am not saying that search is dead - what I am saying is search vendors have a very narrow-minded view of what "search" is all about - they completely ignore the broader aspects of findability that touch on connection mechanisms other than content. I am also arguing that user-generated tags add value to search and create greater findability and re-findability than search against content alone. Tags also acts as broadcast "posts" to a network public which other people can connect to - so there is a non-search domain relationship where tags act as bridging mechanisms to other people with similar interests.

You're argument on defense related agencies lacks merit - I've talked to people in such agencies and tools that enable greater collective intelligence - including tags - are indeed viewed as important trends being monitored.