|SportsML - new language for sports related data|
not sure if this is the right place for this.
News Article [sfgate.com]
Just thought those with sports related web sites might be interested in looking into a new language, SportsML [sportsml.com]. From the article it looks to be XML [w3.org] based following IPTC [iptc.org] standards.
Greetings & Welcome to WebmasterWorld!
SportsML has been in the works since March 2001. The IPTC did, however, just release version 1.0 of the Draft Specification.
|The IPTC launched the SportsML project in March, 2001, as part of a larger effort to track or create specialized vocabularies for data of interest to the news industry. |
In essence, they created the ML to accomodate the reporting of sporting events scores/standings. The IPTC is, after all, a News oriented organization (International Press Telecommunications Council).
I'm currently working on an application which touches on SportsML so it'll be interesting to see how useful their Spec really is.
I was just browsing for any info I could find on the subject, just did't see it mentioned here. So, I thought I might do it for a change...
This will be my first sports related project, so I had not heard of it before. Just started work on the project and we are looking seriously at Sportswire [sportwire.sourceforge.net] using the new SportsML. The language has all the attributes we need and since its xml format is so streamlined, how can we go wrong. (especially with a GNU license)
If you know of a better solution, that would be greatly appreciated. TIA (let me know how your solution works out if ya can)
If only the Lakers would trade S. Walker, getting tired of dunks which turn into rebounds for the other guys... ;)
I posted this URL in another thread but you may find it useful to review:
Basically, it's a list of Projects/Specs for XML based applications.
My application of XML will be for a specific sport and only part of what we're producing will be useful to the News agencies. Our application covers a bunch of other non-News related data as well and that's why I'm a bit worried that SportsML won't meet the need. That being said, however, I know that I can use SportsML to meet the need for what it was intended and create another Schema/Spec for what's not covered.
Just haven't gotten there quite yet!;)
This ML may have arisen out of the amusing (and in some ways scary) fact that their ontology for news contains about 50 times as many terms for sport as it does for all kinds of war, civil unrest and natural disaster.
|...terms for sport as it does for all kinds of war, civil unrest and natural disaster |
Don't they use the same terms (e.g., pound, killed, maimed, destroyed, wiped out, sacked, wounded)!
As the co-author of [sportwire.sourceforge.net...] , what I can tell you about SportsML is that it is the international standard for sports news and stats exchange, and also that there are no vendors of sportsML-based news feed in the North American market. Because Sportwire was required to work with several very different and very broken XML-like sports news feeds, we latched on to SportsML to be the lingua franca of our system -- to support a new feed, all we have to do is to write the input filters to fix the vendor XML problems, then write XSL to translate that feed into the SportsML equivalents. From there, because the system is based on a standard, the website can render or store any of the data that comes across.
Sportwire is far from a finished work; it was developed to suit a specific need, initially for cbc.ca/sports, then later adapted to power the Salt Lake Olympics website for the CBC, and finally to support the best-we-got XML feed from SportsNetwork. That said, I am currently working with SportsNetwork's engineer to streamline the sports news feed: Instead of using 80's technology to ship a 60's technology markup, I'm advocating using RESTful webservice models to ship stats via a simple HTTP service. I just heard from BJ that he's got my first prototype script installed on their server and we'll be checking it out later today. This one small change puts SportsNetwork miles ahead of SportsTicker and AFP in the delivery of the XML data.
As for how to transform into SportsML and the intricacies of the XSL, I'll defer that to my development partner whom I've alerted to this discussion and have asked to poke in here when he's got a minute. Paul is probably North America's leading expert on the complete SportsML spec.
[edited by: DaveAtIFG at 3:12 am (utc) on Nov. 15, 2002]
hello teledyn and welcome to WebmasterWorld!
Very interesting. The sportsite I'm working on will provide XML-RPC as well as RDF/RSS news feeds. The data we're shooting to provide will be applicable to the newswire services as well as those communities which follow our particular sport. I'm interested in learning/exchanging more information on how this sportsML is being implemented and accepted in the news world.
Welcome to WebmasterWorld teledyn, always glad when fellow Canadians show up.
This is a very interesting topic. I have been messing around with hockey scores, because I am in a million pools and wanted to have a central place to track all my stuff. It also gave me a little project to mess around with a few new things.
This thread sparked quite a few hours of research.
>> This thread sparked quite a few hours of research.
sorry, I guess that reads a little funny
just trying to find out about feeds and where to get them and how they are setup etc.
Mostly I found out that it is a vast area and I took down a ton of stuff and have a lot more to do. It pretty much felt like I had identified areas of interest and their connections with each other and some resources, just need to delve into the individual aspects.
>> This thread sparked quite a few hours of research.
which spaked the need for many, many more, as always.
Now you know how I've been feeling for the past few months! :)
There is a ton of information out there and while the concepts are relatively simple, wrapping your mind around the mechanics of the rules and the technology does take a few tries. Especially if you're new to XML. But, just like HTML and PHP and?, it will become old-hat soon enough.
But it does help to have a few mates to bang around ideas with - thanks folks.
Gosh, I hope we didn't scare away OperaMosIEscape
Hi everyone, I'm "North America's leading expert on the complete SportsML spec" :) Lorax, let me know what issues you run into with the spec. Have you joined the SportsML list? We are quite open to feedback since it's still in draft stage with official 1.0 scheduled for approval in March.
If you check out the stats at cbc.ca/sports you can see what Gary and I have done with Sports Network's feed. Like Gary said, we run two transformations: sports network --> sportsml --> html.
Interested in what you're going to do with RSS and XML-RPC.
Welcome to WebmasterWorld [webmasterworld.com] polvo
Welcome to WebmasterWorld.
I'm building a website that is initially targeted at regional sport teams but will be marketed to international teams within 2 years. Single sport. We will be developing xml based stats, offering RDF/RSS news feeds of the games highlights, and offering web-based services through either XML-RPC or SOAP or WDDX (not sure yet). There are several layers to the content we plan to provide beyond simply sports scores. SportsML, from what I've seen so far, will only meet a portion of what we need to produce. But I can't say for sure until we really get into the meat of the project.
WOW, I am amazed. I ask 1 simple question and the true authorities show up. What a place!
Thanks for the link Gregg (lorax), that page is going to eat up all of Sunday just to figure out what I want to read first. Not scared away.. ha ha, just very busy. To give you an idea it's after 1am and just got home from the office, same drill all week... and next.
teledyn & polvo - Thanks for stopping in with your input. I am looking forward to getting started soon, still waiting on hardware though. You both have done an outstanding job and I'm sure (fingers crossed) we are going with the Sportwire/SportML combo. Just so much to do before we think about software, but going with what looks to be the industry standard is a much smarter road than trying to invent a new wheel, usually. I will try to get some feedback to both of you in the future once we get things rolling.
Gregg, I would luv to chat with you about what we are up to, but you know how paranoid the people with the money are.
>>>through either XML-RPC or SOAP or WDDX (not sure yet).
I am sure you know much more than I on the subject, but I would choose the XML-RPC option over the other 2. For simple reasons really, at first glance SOAP seams just a little more complicated to code and WDDX is not a standard and does not look to be anytime soon. I'm kind of a fan of anything W3C has put their stamp of approval on, but thats just me. Buy hey, if you got a square hole, put the square peg in it... use what works the best for your sys.
chat later all, thanks.
|Gregg, I would luv to chat with you about what we are up to, but you know how paranoid the people with the money are. |
You can sticky me anytime.
I agree with you about WDDX. It is in review but there doesn't seem to be any real advances. SOAP appeals to me only because it is purported to have additional flexibility over XML-RPC but at this stage of the game I'm not sure I'll need it. My gut instinct is to stay with the simplest of them all XML-RPC.
Stay in touch. It seems there are only a few of us XML types on this Board so the more we communicate the better.
If I can just throw some caution into the collective sails, I wouldn't wait too long before considering the software itself: It is, afterall, traditionally on of the most expensive parts and you don't want to make it any more expensive than it has to be.
We were recently invited to put together a plan for a sports-gambling site. Unfortunate for them, they were committed to so many bad choices /before/ they called, the cost to dig them out of their hole was outrageous. You might want to touch base with us in our openprojects IRC or bring one of us in as a consultant for just a couple of hours to ensure your planning is in step with the technology track, or at least to illuminate any repercussions of your choices.
As for sports webservices, which I personally believe is largely Emperor's clothes meant to feed consulting companies (every app I've seen for SOAP/.Net could be done more elegantly with plain RESTful HTTP and XML -- budding webservice people are invited to do a google search for the RESTwiki to get tuned into all this), I'll stay any final judgement until I know more about the task, but for the sports-related webservices I've pondered, things like instant updates, headlines, scores, odds &c, SportsML is overkill
SportsML is intended as a storage/transmission format for rich multimedia content such as is seen on news websites or even news broadcasts; it might make sense if you are wanting to present your sports news website over WAP, but for simple score and stat updates, it makes more sense to work from RDF with an appropriate sports-specific namespace. I think SportsML would inflict a lot of overhead, and with RDF you can leverage the tools and development in the RDF world (for an extreme example, the data mining experiments of [mozilla.org...]
Just a last note on webservices, you may be interested to know that it was a webservice that brought me here! On the Sportwire.sourceforge.net website, we have a sidebar that lists referrers, and after the NYC announcement, I checked into the site to see if we were being cited ... lo and behold, there was the backtrack link to here! :)