Comment Re:In case you have no clue what they're talking a (Score 1) 135
What does the Semantic Web offer, why do we care?
I'm going to try to add some personal perspective in addition to the worthy Wikipedia article linked in the parent, because I see a lot of criticism in these threads and not for the traditionally criticism-worthy issues. In case you're wondering, I was involved in a non-trivial Semantic Web related project in 2005: a learning experience, I won't mention it further. That said, I could be _totally_ wrong; but this is how I see things.
The first question: Why do we want the Semantic Web? Sure, it sounds fancy, but why should _you_ (the average Slashdot reader) be excited about it? Well, let me explain why _I'm_ excited and maybe you'll agree... I tend to move data around, through systems and people, changing formats as necessary, making logical decisions based on the data as appropriate. Often there's a convenient library or tool or API for assisting me in doing this, making my life easier by abstracting the process of getting at that data and mashing it around into something more immediately useful. From my perspective, the Semantic Web will give me that power at a new level of convenience. Semantic markup, formats, ontologies, etc, allow data-centric code to be written more quickly and with less reinventing of wheels. Ever written a screen scraper? A perl script to pull data out of a proprietary log format? The Semantic Web will not be a panacea for these kinds of problems, but if we can convince people to mark up more data in reasonably common/standard ways then hopefully things like software mashups should become easier than ever! When software is better able to understand what data _is_, without huge amounts of domain-specific programmer effort, making decisions based on that data should be easier. Take a look at Firefox, microformats, and SPARQL, for example. Do users care about the Semantic Web? I don't think so, because all they should see is basically the same old browser-rendered Web. However, our ability (as software developers and general geeks) to produce useful tools and websites using Semantic Web data may result in even better websites and dynamic services.
Second question: What does the Semantic Web look like? Not like Gravity, in my opinion. Dynamic graphs are handy visualization tools for some kinds of data, but definitely not all! In fact, they're pretty brittle and they don't scale at all. There are a lot of interesting proposed solutions to the visualization problem (see SIMILE and MIT's Haystack), but I don't think it really matters. Within a specific domain, there will always be better visualization tools than a generalized visualization method (written by those familiar with the domain). So, the Semantic Web will look basically the same as the current web. In fact, if you start looking carefully, I think you'll see it all around you...
Third question: Why is "open data" exciting and what's the difference between just opening a MySQL database to the public and the Semantic Web vision? Well, if a site is exposing its "database" in RDF using a common ontology, then you can make use of their data just as you'd use their services via an API. A data provider may not foresee all the potentially useful ways to use their data just as they may not foresee ways to make use of their API, but a clever programmer can take from their surroundings what is needed and make of it something more. If you think this is random, note that /. and k5 have been serving up RDF of their frontpages for years and that today we regularly use RSS feeds and some black magic to do similar things.
As I said, I could be way off the mark here. This is just the simplified perspective I've adopted after thinking about it for a while and reading the common sources. Please don't take this as gospel or thorough, comments or corrections are very welcome.
I'm going to try to add some personal perspective in addition to the worthy Wikipedia article linked in the parent, because I see a lot of criticism in these threads and not for the traditionally criticism-worthy issues. In case you're wondering, I was involved in a non-trivial Semantic Web related project in 2005: a learning experience, I won't mention it further. That said, I could be _totally_ wrong; but this is how I see things.
The first question: Why do we want the Semantic Web? Sure, it sounds fancy, but why should _you_ (the average Slashdot reader) be excited about it? Well, let me explain why _I'm_ excited and maybe you'll agree... I tend to move data around, through systems and people, changing formats as necessary, making logical decisions based on the data as appropriate. Often there's a convenient library or tool or API for assisting me in doing this, making my life easier by abstracting the process of getting at that data and mashing it around into something more immediately useful. From my perspective, the Semantic Web will give me that power at a new level of convenience. Semantic markup, formats, ontologies, etc, allow data-centric code to be written more quickly and with less reinventing of wheels. Ever written a screen scraper? A perl script to pull data out of a proprietary log format? The Semantic Web will not be a panacea for these kinds of problems, but if we can convince people to mark up more data in reasonably common/standard ways then hopefully things like software mashups should become easier than ever! When software is better able to understand what data _is_, without huge amounts of domain-specific programmer effort, making decisions based on that data should be easier. Take a look at Firefox, microformats, and SPARQL, for example. Do users care about the Semantic Web? I don't think so, because all they should see is basically the same old browser-rendered Web. However, our ability (as software developers and general geeks) to produce useful tools and websites using Semantic Web data may result in even better websites and dynamic services.
Second question: What does the Semantic Web look like? Not like Gravity, in my opinion. Dynamic graphs are handy visualization tools for some kinds of data, but definitely not all! In fact, they're pretty brittle and they don't scale at all. There are a lot of interesting proposed solutions to the visualization problem (see SIMILE and MIT's Haystack), but I don't think it really matters. Within a specific domain, there will always be better visualization tools than a generalized visualization method (written by those familiar with the domain). So, the Semantic Web will look basically the same as the current web. In fact, if you start looking carefully, I think you'll see it all around you...
Third question: Why is "open data" exciting and what's the difference between just opening a MySQL database to the public and the Semantic Web vision? Well, if a site is exposing its "database" in RDF using a common ontology, then you can make use of their data just as you'd use their services via an API. A data provider may not foresee all the potentially useful ways to use their data just as they may not foresee ways to make use of their API, but a clever programmer can take from their surroundings what is needed and make of it something more. If you think this is random, note that
As I said, I could be way off the mark here. This is just the simplified perspective I've adopted after thinking about it for a while and reading the common sources. Please don't take this as gospel or thorough, comments or corrections are very welcome.
