If you want to make the case for "improved" searching via the wonders of semantic-web technology, as this blog post and this demo attempt to, you need to make your demo demonstrate something compelling.
In the blog post announcing that demo, Kingsley suggests "Microsoft" as the query example. As of this moment, doing that query on that demo produces a page of unintelligibly elided URLs, misrendered characters, and random blobs of text that contain the word "Microsoft" in them. The UI opens with this stirring invocation:
And the first search result begins, and I feel like I have to clarify that I am not making this up, with the words "Mac OS X Leopard" (and then some gibberish that I'm guessing used to be Italian).
If you do the search "Microsoft" on Google right now, you get some news items about Microsoft, followed by the Microsoft site itself.
But maybe that was just an unfortunate example. So I tried looking for Cyndi Lauper. Google's results for this begin with Cyndi's official site, then the Wikipedia page about her, then her MySpace page. OpenLink's begin with "The Parking Lot 03.09.2007 at SmartLemming.com", again in a page-layout that isn't even funny as a parody of good information design.
If you want to amuse yourself by trying more examples, I've put up an easy form for running a search on both sites side-by-side:
cyndi lauper
microsoft
(try your own)
Be patient with the OpenLink side...
To state the obvious caveat, the claim OpenLink is making about this demo is not that it delivers better search-term relevance, therefore the ranking of searching results is not the main criteria on which it is intended to be assessed.
On the other hand, one of the things they are bragging about is that their server will automatically cut off long-running queries. So how do you like your first page of results?
And on the other other hand, the big claim OpenLink is making about this demo is that the aggregate experience of using it is better than the aggregate experience of using "traditional" search. So go ahead, use it. If you can.
Now, did your opinion of the potential of the "semantic web" go up or down during your experience?
[Update: Kingsley responds here, and suggests that "glenn mcdonald" would actually be a better example query. So here you go: glenn mcdonald. Did your opinion change?
Just to be clear, I think Kingsley is exactly right that we need a universal data browser, and quite possibly right that Virtuoso's underlying technology is capable of being an engine for such a thing. But this thing he's showing isn't a data browser, it's a data-representation browser. It's as if the first web-browser only did View Source. We will no more sell the new web by showing people URIs than we sold the old web by showing them hrefs. Exactly the opposite: we sold the old web by not showing people UL and OLs and TD/TD/TD/TD and CELLPADDING=0. And we'll sell this new web by not showing them meta-schema and triples and reification and inverse-link entailment.]
In the blog post announcing that demo, Kingsley suggests "Microsoft" as the query example. As of this moment, doing that query on that demo produces a page of unintelligibly elided URLs, misrendered characters, and random blobs of text that contain the word "Microsoft" in them. The UI opens with this stirring invocation:
Displaying values and text summaries associated with pattern: (NULL)1
(NULL)1 contains "microsoft" in any property value.
(NULL)1 contains "microsoft" in any property value.
And the first search result begins, and I feel like I have to clarify that I am not making this up, with the words "Mac OS X Leopard" (and then some gibberish that I'm guessing used to be Italian).
If you do the search "Microsoft" on Google right now, you get some news items about Microsoft, followed by the Microsoft site itself.
But maybe that was just an unfortunate example. So I tried looking for Cyndi Lauper. Google's results for this begin with Cyndi's official site, then the Wikipedia page about her, then her MySpace page. OpenLink's begin with "The Parking Lot 03.09.2007 at SmartLemming.com", again in a page-layout that isn't even funny as a parody of good information design.
If you want to amuse yourself by trying more examples, I've put up an easy form for running a search on both sites side-by-side:
cyndi lauper
microsoft
(try your own)
Be patient with the OpenLink side...
To state the obvious caveat, the claim OpenLink is making about this demo is not that it delivers better search-term relevance, therefore the ranking of searching results is not the main criteria on which it is intended to be assessed.
On the other hand, one of the things they are bragging about is that their server will automatically cut off long-running queries. So how do you like your first page of results?
And on the other other hand, the big claim OpenLink is making about this demo is that the aggregate experience of using it is better than the aggregate experience of using "traditional" search. So go ahead, use it. If you can.
Now, did your opinion of the potential of the "semantic web" go up or down during your experience?
[Update: Kingsley responds here, and suggests that "glenn mcdonald" would actually be a better example query. So here you go: glenn mcdonald. Did your opinion change?
Just to be clear, I think Kingsley is exactly right that we need a universal data browser, and quite possibly right that Virtuoso's underlying technology is capable of being an engine for such a thing. But this thing he's showing isn't a data browser, it's a data-representation browser. It's as if the first web-browser only did View Source. We will no more sell the new web by showing people URIs than we sold the old web by showing them hrefs. Exactly the opposite: we sold the old web by not showing people UL and OLs and TD/TD/TD/TD and CELLPADDING=0. And we'll sell this new web by not showing them meta-schema and triples and reification and inverse-link entailment.]