Staff researcher David François Huynh has created an interesting tool for browsing semantic database Freebase, called Freebase Parallax. Written up by ZDNet's Oliver Marks, the video Huynh recorded demonstrating Parallax (below) will knock your socks off.
Unfortunately, actually using Parallax demonstrates just how far from solid Freebase, one of the semantic web's poster children, really is. The idea is to allow you to apply multiple filters for your searches and embed live charts in a blog. It's a beautiful idea, check out the video.
Here's the video below, if you find yourself saying "get to the point already," then skip to about 1:30 in the timeline.
Unfortunately, when we tried out a number of searches in Parallax, very few subjects were well populated at all. We found duplicate subject titles where one held solid data and the other didn't, but even that was a best case scenario. In search after search, we found next to nothing in Freebase.
The example above is nice, but let's say I want to find out something about black women scientists. No luck. History of the internet? Not much information there. Venture Capitalists? Blank profile pages.
This ought to work. Freebase has taken more than $50 million in venture investments, they have a small army of volunteer and computer scientist contributors, they've got robots pumping their database with information automatically. There are now 60% more articles in Freebase than there are in English Wikipedia. So what's the problem?
We wrote last week about ontological concerns about the semantic web, but Parallax shows that there are more superficial problems. An unfriendly UI has been Freebase's excuse for a long time, despite recent improvements to it. We love the idea of the semantic web, but give it's grand daddy website a usable UI like Parallax and we're left questioning just how much there really is inside Freebase anyway.
For an alternate view see Alex Iskold's Freebase: Dispelling the Skepticism, and some fault here may lay in the coolness ratio of the video to the Parallax app, but for now - we feel inclined to look elsewhere for the "semantic web killer app."
Disclosure: The author has consulting relationships with a number of pre-launched semantic web companies.
Comments
Subscribe to comments for this post OR Subscribe to comments for all Read/WriteWeb posts
Hi Marshall,
I guess the question is... how good is good enough?
For the semantic web ( freebase's subset of it, actually ) to answer your questions, a lot of data needs to be in place.
Freebase is some sort of a centralized effort to build this database. And that my friend, takes time.
However, I think the true value will come from the other side, from grassroots adoption.
The killer app is already coming from that side, and it is called Linked Data.
If RSS created a complete information revolution, imagine what will happen with a little more RDF scattered around the world. Around the enterprise.
It's a complex system.
I think you should cover Sindice next in the SW saga, and then point people to Kingsley Idehen's Linked Data Browser ( OpenLink ).
Nice to see David's remarkable work mentioned here... ;)
Best,
A
Posted by: Aldo Bucchi
|
August 13, 2008 9:11 PM
@Aldo, you've correctly identified the problem with Semantic Web, but you're incredibly optimistic in your hopes for a solution. I don't see it taking off because it's just more work for content creators to do with little benefit.
Posted by: Jason Carreira
|
August 13, 2008 9:17 PM
@Jason
Yes, there is little incentive to publish. But that's not true for ALL publishers. Just those who are counting on capitalizing on the attention generated at the consumption of the data. ( like RWW ).
What about those who need to publish their data for human users, and then need to create APIs to deliver the same data in a machine readable form?
That's an incentive right there, and a BIG one. I have been using RDF for five years to do just that. As middleware in the enterprise, and it pays of.
The problem in that case is more about skills and tools. But that is changing too.
We will soon have our identity + data trying to merge through the walls of many internet properties. RDF is a smart, cheap, standard choice.
Let's say that I agree with you, but I am optimistic because I know how really powerful RDF can be from the "data bus" and blackboard perspective, and I am counting on more people realizing that in the short term. ( it's all in the power of the URL ).
The SW is already here. But it's gon
Posted by: Aldo Bucchi
|
August 13, 2008 9:30 PM
Freebase, huh?
Posted by: steveballmer | August 13, 2008 9:42 PM
Liking it cause I think the life science informatics people will have some opinions. IMO, Parallax is an amazing example of what's possible with structured data (and there are a number of example that Freebase has done a great job of highlighting, even if it doesn't quite fit optimally onto the semantic web)
Posted by: Deepak
|
August 13, 2008 9:49 PM
Marshall, I'm a long-time fan of David Huynh's work, but I have to agree with you about how the richness of his interface reveals the poverty of the Freebase data. Maybe it's just a matter of time or money, but Metaweb seems to have had enough of both. I'm still waiting to see them build something that lives up to their logo: an "open database of the world’s information."
Posted by: Daniel Tunkelang | August 13, 2008 9:55 PM
http://www.w3.org/TR/WD-rdf-syntax-971002/ vs. http://groups.google.com/group/comp.lang.java/browse_thread/thread/6923c024ed392c85/88fa10845061c8ba - So Google got less than a year's head start on the SW. 11 years (!!) later, the SW is still gonna be useful, really! Any day now.... (Seriously - some of the stuff out of SIMILE was useful on the real-web. But point me at any other working, non-demoware SW site...)
Posted by: Nick Lothian
|
August 14, 2008 12:02 AM
Disappointment has to be taken very seriously and reasons have to be analyzed. For instance I am disappointed of Twine.com, too. Some millions have been invested, so what are the causes for underperformance? Not the over expectations.
To make some new friends now ... I really like you guys out there, but there is still too much tech thinking dominating in these keen projects, There is not enough filtering the priorities with hmmm business criteria, I suppose. My prejudice? I am living in Europe and to make people "think in business terms" is more difficult here.
That simply means to analyze the value and behavior patterns of the user seriously and make UI issues for instance a top priority during the whole process.
The image of SW in the public is still rather academic. To give the "movement" a new momentum I propose to speak about "semantic business" instead of "SW" in some contexts to make anyone listen again.
Posted by: Willi Schroll | August 14, 2008 2:33 AM
Will, you make a good point. At Endeca, we don't position ourselves as a "semantic web" company, even though our technology would certainly allow us to do so. Instead, we focus on creating business value and bringing in revenue.
Posted by: Daniel Tunkelang | August 14, 2008 5:52 AM
Maybe there are many domains that are lacking, but this is still one of the best and inspiring demonstrations of what can be done with nicely organized data.
Still, this is general-purpose tool and it is actually complicated to use. What I would like to see is very special-purpose tools that can run circles around the tools that try to do similar things without organized data.
Hope to see more from Freebase, the guys there really are doing something worthwhile!
bye
Andraz Tori, Zemanta
Posted by: Andraz Tori | August 14, 2008 6:51 AM
@Nick: Oh, ... semantic web useful application - Zemanta is partly built on semantic web data. And it's non-demoware.
@Will: IMHO in Europe Semantic web is slow because too much energy is being poured into SW academic research instead of applying it to business.
Posted by: Andraz Tori | August 14, 2008 6:56 AM
Re commercial semantic web applications in Europe: how about Garlik?
Posted by: Daniel Tunkelang | August 14, 2008 10:34 AM
Forget about SW or non SW labels.
Which technology companies call themselves HTML companies?
not many ( at least not today )... they focus on their value proposition.
In the end its all about the standards with which you open your info to the world that will make you a SW company.
Endeca could choose to publish their data using SW standards, and that would add to the SW.
Will they?
If it becomes a business to do so, they will.
If customers ask for it, they will.
If RDF tools mature enough that they are cheaper than inhouse solutions, they will ( perhaps internally at first, but this will eventually surface at some points ).
Just don't expect this to happen overnight.
Posted by: Aldo Bucchi | August 14, 2008 10:43 AM
Aldo, that's a good point. At Endeca, it's actually our customers who would be deciding how to publish their data, but you're right that their motivation will have to be that it creates business value for them.
Posted by: Daniel Tunkelang
|
August 15, 2008 5:23 AM