More Than Who You Know, It's What You Know

There is a nice little article in the Economist this week talking about the glimmering appearance of semantics in web applications. We are seeing a blurring of Web 2.0 and semantic web ideas [old ppt download]. Web 2.0 is a poor man’s semantic web. I think the poor man's view is about to start getting richer.

The real value-added of semantics struck me as I was (finally) playing with Facebook a few months ago.  I’ve been on LinkedIn for years now, but haven’t otherwise been drawn to social networking sites.  Facebook has become such a force of nature, it has been hard to resist invitations to “become friends” by people I already consider friends, for fear of otherwise denying our real world friendship.  As I became engaged with Facebook I soon realized what’s there, and more importantly, what’s missing. 

Facebook as Peopleweb Lite
Facebook is a schema and some communications methods.  It defines certain data types (people, events, messages) and their attributes.  So it provides a simple set of semantics to allow me to declare a person (name, address, etc.) and the form of relationship (worked with, dated, etc.) The user explicitly declares everything. Nothing is inferred.

The brilliance of Facebook lies in its simplicity and its completeness.  There has been a standard around for years now in the semantic web community called FOAF – for Friend of a Friend.  FOAF is an RDF and XML standard for defining a person’s attributes to make a uniform, web-distributed method of inferring contact information. FOAF has never risen past the status of demonstration project, even though the technology is quite simple.

Like FOAF, Facebook has its own standard description of a “person,” but addressed the additional issues of privacy (by controls) and discovery (by directory).  Bingo! A proprietary standard wrapped in a walled garden becomes useful.  Then the garden gets a gate (Facebook API) and now it is called an ecosystem.

The Paradox of Simplicity in Social Network Design
But even with this simple semantic web for people, Facebook lacks real utility beyond entertainment, as do most social network sites.  Why?  Because they are defined around fixed schemas (schemata for you purists) which are the boundaries of the community.  LinkedIn is a fixed schema around resume history.  Myspace is a fixed schema around people and entertainment media. 

Social networks are a subset of the general class of community websites. The essential value-add of a community site is to define the schema of the community – what we have in common as members of the community become the defining features of the site.  The granddaddy of community sites is Ebay.  What is Ebay but a schema (marketplace) and a set of rules for interacting around the schema?  What is Amazon but a schema (product catalog) and a set of rules for interaction?  This simple rigidity of design makes these sites intuitively obvious to the casual user and inherently bounded in their ability to evolve past the founding metaphor.  Again, as the textbook example of a self-limiting metaphor I would point to Ebay.

I find social networks of limited utility, because they only do one thing, albeit well.  They connect rigid entities (usually people) for a limited purpose (discovery or communication).  They are shared contact databases with access control and messaging. While they are file sharing containers, they are not repositories for knowledge.  The knowledge inherent in my community is not directly accessible, only addresses. Social networks as presented today don’t help us actually do anything together.

From Social Networks to Knowledge Networks
A social network resembles the old telephone party line. While it connects us, it adds no value to the conversation.  It is the intermediary, momentarily connecting the knowledge that remains mostly in our heads. All the real action still happens at the edge of the social network, not in the network.  They provide no long-term memory or framework for accumulating and navigating what we collectively know.  I cannot imagine using Facebook to discover, accumulate, and share diligence on a prospective investment beyond figuring out who to call for an indirect reference.  Facebook's peopleweb doesn't contemplate how I might accumulate information about a person, their company,  its competitors, and customers and then tap into the related knowledge that my partners or business acquaintances might have.  Yet, the diligence process is one of connecting dots of knowledge (conversations, emails, documents, relationships) to find a pattern that yields an outcome.  The same is true for many collective activities in our communities and organizations. Hence the varieties of definitions around community - communities of purpose, communities of practice, etc.  Communities arise from common circumstance, but they persist because of a shared utility of being.

This is why I think social networks will evolve into knowledge networks.  All the pieces are place.  Facebook has shown us how a gated community can encourage declaration of knowledge (about people and relationships) and maintain a sense of control and privacy.  What needs to be extended is the method to for defining arbitrary entities and concepts beyond people (companies, locations, events, bands, diseases, hobbies, TV shows, chemicals, etc.) and allowing users to define relationships between them (owner of, owned by, included in, compatible with, etc.)  What I have just described is an explosion of meta-data is far bigger than the web itself. But the path is now clear.  No AI.  No inference engines, No voodoo.  “Just” a very large and extensible graph.

We won’t get there in one big leap.  But we will get there, one extension at a time.  Today’s social networks are a transitional Web 2.0 concept.  As they add more and more semantics about different types of objects and concepts, they will morph from social networks to knowledge networks. And knowledge networks will be more valuable and stickier, because what you know is often more valuable than who you know.

Postscript
Well, it seems the spooks have already come to the same conclusion (NYT article) in designing "A-space." They need a social network for the knowledge they gather in espionage and counter-espionage. All the more reason you'll need one soon, too.

Comments

[Erkko]

You really put it in words that I have been thinking of for a while. Even had a draft post about this waiting for further writing, but this really summed up the limitation i've been feeling with our social networks, but just before haven't been able to really pinpoint it down.

Brilliant!

[anthropocentric]

Geni is an interesting combination of a social network / almost-wiki. While it facilities communication, it is primarily a tool to help build and develop a family tree.

[DisMonkey]

A great post! Knowledge networks are the way of the future. Information is no longer static or sparse, and the tools that manage information need to catch up with that.

I'm sending you an email with some further thoughts you might find interesting.

-Nick

[Michelle Bonat]

Thanks Peter your post made my day! It is delightful to see a true understanding of the impact of the sem web. The follow on discussion is what are the next proof points or "extensions" - there are boundless opportunities in bringing this technology to specific audiences to solve pressing problems. Love the term knowledge networks...mind if we borrow it? - Michelle

[Jason Vu]

We will see which social networks will evolve into knowledge networks. Can Facebook make it?

[Don Jones]

Knowledge networks may be more challenging due to human nature. It is easy to participate in a social network, the cost is relatively low because I'm just telling you some things about me - my likes, dislikes, etc.

When it comes sharing my knowledge, that becomes more labor intensive and I also perceive some of my knowledge to have greater value, so I may not willingly share it without some form of compensation...Therein lies a greater friction...

I'm sure some venture capital backed company will figure it out - maybe "Pay Per Idea!"

[Sin Jin Lee]

It seems that the major social networks operate on exchanging surface level information. Pictures and videos and comments. But to take the info that is already there-occupation, relationship status, etc and to extrapolate from that data and offer users new permutations of that info is def exciting.

Word on the street is that Radar Networks might be harbingers of this Web 3.o like world. Potential app: you want to have dinner this Friday night in San Francisco. A web 3.0 app fishes through all of your contacts on Facebook, finds out which one's will be free for dinner, email them invites, and sends out a reservation to the Restaurant that everyone wants to eat at.

The extent to how much "knowledge" will actually be willingly shared on these knowledge networks is a fuzzy future picture. With so much individual wrangling over who owns what-who gets compensated for what ip, patents, copyrights-I feel like a fluidly accessible knowledge network might be quite a few years away.

[Shai Berger]

Great post. Three things...

1) "What is Ebay but a schema (marketplace) and a set of rules for interacting around the schema? What is Amazon but a schema (product catalog) and a set of rules for interaction?"

Replace "schema" with "database" and these statements aren't as exciting. What is *any* website but a database with a html based front-end?

2) "This simple rigidity of design makes these sites intuitively obvious to the casual user and inherently bounded in their ability to evolve past the founding metaphor."

I think the "rigidity of design" of a social network is actually its *key* defining factor....

* MySpace is too flexible. You start with a blank page and then, faced with so many options, many people found it daunting. Since pages are so "free form", the platform lacks a sense of cohesion.

* Linked-In is too rigid. You just "fill in the blanks". That makes it very comfortable for people to join (the barrier is very low) but the pages are lifeless.

* Facebook is just right. You fill in the blanks to get started and right away you get a page that looks complete and "alive", yet you still have many options for adding and configuring.

People don't give enough credit to the million little decisions that create the rigidity/flexibility balance in a social network, and how critical that balance is to success.

3) "Yet, the diligence process is one of connecting dots of knowledge... to find a pattern that yields an outcome."

In my view, that pattern finding is a unique talent of the human mind. It's the magical step in converting data into knowledge. Communication technology only lets us automate more and more of the effort around that process. But that's still a great thing - the more I automate the mundane, the more cycles I have left for critical thinking. Every little bit counts. Outlook auto-complete's email addresses... great that's 3 seconds saved. Linked-In reminds me from where I know someone, etc. The emergence of semantic web is another step in that direction, but diligence research will still require the brain of a VC -- or, more likely, an overworked intern. :-)

- Shai

http://www.shaiberger.com

[Mike Walsh]

Thanks for your post, Peter. We are actually beginning to witness this happen within many of our clients' social networks. Through tagging, visualization, relationship mapping, and ratings it is becoming easier to identify the appropriate knowledge and people assets along with the sources and paths of those assets. This will become common-place within companies into 2008.

[Leobard]

The web 2.0 is getting richer based on the proprietary walls.

You mix up cause and effect: FOAF as a standard is making Web 2.0 services interoperable and standardized. Orkut, facebook, linkedin, studivz, (++) would all have the same API and an extensible data format if they used FOAF and RDF.

But wait - who is the venture capitalist behind these web 2.0 walls? Maybe you. So you suddenly realize that when your little nerds in the computer room switch the lever towards standards, your money may go down the well, because then your precious closed community of people, and that is what you sell and own: data about people, will be open for anyone else to copy. So you would do your best not going for standards but instead making the BEST social service EVER so that everyone DIGGS it and invites all HIS FRIENDS into closed walls. Capitalism is ok, name it what it is.

Microsoft does not go crazy for a standardized OpenDocument format, which would help us get free from the monopoly that takes your money when buying a computer and invests it to sell XBoxes to your kid. Why should it be different with web 2.0? But thanks for pointing us to the VC view.

[Yaron Bazaz]

Hi

I beliave we can look at the internet as an evolution of neurological system.

Starting as a thin network of a simple and isolated "neurons" (meaning - web sites, or servers), with almost no connections among them (resemble some primitive forms of life) and continue to the current web stage – a mixture of standalones web sites together with initial formations of social networking/info sharing sites.

I predict that the internet will become a huge, fully integrated, learning and acting neurological system. Have a look at a fully developed neurological system like the human brain as an example. It is not a primitive network of independent cells anymore; it’s the most sophisticate's organ in the known universe. Each neuron has connections to thousand others neurons. The aggregative impact of "data" release (discharge patterns of action potentials) from huge amount of cells, (while each neuron is still an autonomies cell) in a specific brain area/ neural circuit produce meta - outputs like actions, thoughts, body function maintenance, feelings and more. The web will be much greater than the sum of his parts. Each server (or web site) will be connected directly to ten thousands other servers/sites, and the aggregate information gathered from a given chunk (organized around categories or tags) will influence much higher hierarchies, "control" the way the world act, react, feel and maintain our life.

Regards - Yaron

[Susan Sharma]

Your post has helped me conceptualise where my own community online "IndianWildlifeClub.com' can go. We have tried to be a knowledge network for wildlife lovers of India. The response from the websurfer is good-but how does one survive and scale up? Paid members is a possibility but few people are ready to pay for knowledge. unless there is a firm possibility of future revenues one cannot get VCs interested.

[Gene Leybzon]

For your point on schema that is capable of reflecting real-wold semantics, I wonder if you came across FreeBase. If it (or something like this) will become de-facto standard for capturing and storing meaningful data and relationship, it could be a unified language for sharing data between different social networks and much more...