Award categories with gold/silver/bronze designations

937 views
Skip to first unread message

Jeff Prucher

unread,
Oct 25, 2013, 6:37:20 PM10/25/13
to freebase...@googlegroups.com
I'd love to get people's opinions about this. I've had a request to find a way to model awards that have multiple categories of achievement. For example, the Clios: http://www.clioawards.com/catalog/2013/branded_entertainment/. Note that they have separate "gold", "silver," and "bronze" levels of award in each category. Currently we would model those as something like "Clio award for branded entertainment in film - Gold", "Clio award for branded entertainment in film - Silver", etc. Another example is the Newbery Medal -- we currently model the Newbery Medal and Newbery Honor as two different categories. 

The other way to do this would be to add a new property to /award/award_honor, something like "level of achievement". That would hold values (probably an enum list) of things like "gold", "silver", "bronze".

This is appealing for a few reasons:
1. The achievement levels don't really represent different award categories: You cannot be nominated for a "Clio award for branded entertainment in film - bronze" or a Newbery honor.
2. Having to create the gold/silver/bronze/etc. levels as award categories creates a huge number of extra topics that can be hard to manage (there are something like 95 current Clio categories, and even more historically). 
3. Umm. OK, that was only two reasons.

The major drawback I see to this is loose semantics. What entity would we use to represent the "honor" level for the Newbery and Caldecott awards? And we'd want to try to limit this to actually named achievement levels. We really don't want this to balloon out to runners-up who didn't actually win anything (we call them "award losers" and don't want to . "Honorable Mention" would probably be a legitimate use of this, too. 

Further caution would have to be undertaken, since some awards have various "gold", "silver", etc. designations which don't actually reflect different levels of achievement, it's just that different categories have different names (the gold and silver bears of the Berlin Film Festival, for example).

What do you think?

Jeff

Tom Morris

unread,
Oct 26, 2013, 9:28:55 AM10/26/13
to freebase...@googlegroups.com
I don't find reducing the number of award topics to be a particularly compelling reason, but if the proposal were to include semantics which allowed one to tell which award level was higher/lower, I'd be more interested.  A simple way to do this might just be to use the existing structure and add a property for Higher Award Category/Lower Award Category.  Of course, that doesn't address the Clio award explosion issue.

Tom


--
You received this message because you are subscribed to the Google Groups "Freebase Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to freebase-discu...@googlegroups.com.
To post to this group, send email to freebase...@googlegroups.com.
Visit this group at http://groups.google.com/group/freebase-discuss.
For more options, visit https://groups.google.com/groups/opt_out.

Thad Guidry

unread,
Oct 26, 2013, 11:35:15 AM10/26/13
to Freebase Discuss
Hmm... levels....

Saying the words "gold", "silver", "bronze" means "1st place", "2nd place", "3rd place", etc... runners-up, honorable mentions.

As Tom points out that is more valuable to know that Person X had 1st place in Astrology compared to Person Y and is regarded by judges as being "better" somehow or having a higher importance in the field than the other guy/gal.

Charles Babbage received Gold (1st Place) in 1824 by the Royal Astronomical Society...but who was 2nd Place or achieved Silver ?

1824
Charles Babbage     J.F. Encke
Charles Rumker*     J.L Pons*

The ones with * after the name are the 2nd placers.. Silver medal of the Royal Astronomical Society, which stopped being given out in 1827.

It would also be useful to know when the particular level of achievement was established and then discontinued... just like Award Category has....so...

Maybe just have a Parent / Child property on Award Category to hold "Higher Award" , "Lower Award" as Tom suggested might be best ?

And the "level of achievement" enumerated property added to Award Honor....but there are some weird outliers that should be noted on this property...such as:

( Unusual instances in award history where a "ALL ARE EQUAL PLACE" mentality was affect by judges.  In 1847 when the discovery of no less than 3 planets in our solar system was made by many members of the Royal Astronomical Society, the society decided to have no award at all, and then unpack a "Testimonal" in red vellum to many members during the following year 1848... saying that those particular listed members each had equal importance and status in Astronomy for their contributions during the previous year.

References:


Finally,  I have an interest in seeing all the "2nd placers", the underdogs that did not quite get fame.  How can we easily expose that for users to query ?

Set that up in Schema Staging... we can hack on some instances like the above examples.

-- 

Jeff Prucher

unread,
Oct 28, 2013, 1:39:13 PM10/28/13
to freebase...@googlegroups.com
I like the higher/lower property. It seems like whichever approach we decide on it's something we should have.

I don't really object to the Clio explosion, if that's the right approach. The main reason I'm attracted to the extra property is that, semantically, "Newbery Medal" and "Newbery Honor", or "Clio award for X - Gold" and "Clio award of X - Silver", are not really different categories. And therefore, how would we enter the nominees? You can't be nominated for a "Clio award for X Gold" -- you're nominated for a "Clio award for X", and you can win a gold, silver, bronze, or nothing at all. If someone wanted to compare nominees to winners (say, what percentage of female vs. male nominees for a given award actually won over time?), I'm not sure it could be done unless we have this new property. 

Jeff Prucher

unread,
Oct 28, 2013, 2:10:29 PM10/28/13
to freebase...@googlegroups.com
On Sat, Oct 26, 2013 at 8:35 AM, Thad Guidry <thadg...@gmail.com> wrote:
Hmm... levels....

Saying the words "gold", "silver", "bronze" means "1st place", "2nd place", "3rd place", etc... runners-up, honorable mentions.

It sort of does and sort of doesn't. In most awards, runners-up are just highly-placed award losers. If you look at the Clio website, it's also clearly not a 1st, 2nd, 3rd designation (although for something like the IPPYs it arguably is) -- in a normal competition, if there is a three-way tie for first, there are no second and third-place finishers. But just in the film category, there are 14 gold winners (even if you fragment it further based on sub-category and single-vs.-campaign, there are up to five), and still room for plenty of silver and bronzes.

 

As Tom points out that is more valuable to know that Person X had 1st place in Astrology compared to Person Y and is regarded by judges as being "better" somehow or having a higher importance in the field than the other guy/gal.

Charles Babbage received Gold (1st Place) in 1824 by the Royal Astronomical Society...but who was 2nd Place or achieved Silver ?

1824
Charles Babbage     J.F. Encke
Charles Rumker*     J.L Pons*

The ones with * after the name are the 2nd placers.. Silver medal of the Royal Astronomical Society, which stopped being given out in 1827.

But "Silver medal of the Royal Astronomical Society" is not just a generic second place -- it's an actual awarded honor. 
 

It would also be useful to know when the particular level of achievement was established and then discontinued... just like Award Category has....so...

Hmm... that suggests that, if we did create an "award achievement level" type, we would want award-specific levels, and not just generic gold/silver/bronze. That would allow start/end dates, and also make the higher/lower properties work better. 
 

Maybe just have a Parent / Child property on Award Category to hold "Higher Award" , "Lower Award" as Tom suggested might be best ?

And the "level of achievement" enumerated property added to Award Honor....but there are some weird outliers that should be noted on this property...such as:

( Unusual instances in award history where a "ALL ARE EQUAL PLACE" mentality was affect by judges.  In 1847 when the discovery of no less than 3 planets in our solar system was made by many members of the Royal Astronomical Society, the society decided to have no award at all, and then unpack a "Testimonal" in red vellum to many members during the following year 1848... saying that those particular listed members each had equal importance and status in Astronomy for their contributions during the previous year.

References:


Finally,  I have an interest in seeing all the "2nd placers", the underdogs that did not quite get fame.  How can we easily expose that for users to query ?

If we really wanted to be able to show, for awards that publish non-winners in a ranked order, we might need yet another CVT (parallel to award_honor). The reason I say that, rather than trying to decorate the /award/award_honor CVT with another property, is that the various "award winner" types and properties (award_winner, award_winning_work, etc.) are semantically clear that someone or something won an award. If we added a "rank" property (a la /award/ranked_list), we would lose that clear semantic meaning. Users wanting just winners would have to filter out ranks higher than 1 (but include 1 and null), which I think would be a bad plan.

I'll stick some stuff in /base/schemastaging, and we can see what it all looks like.

Jeff 

Jeff Prucher

unread,
Oct 28, 2013, 7:38:29 PM10/28/13
to freebase...@googlegroups.com
I made a schema on sandbox, since trying to extend a CVT on /base/schemastaging was going to be too hideous to contemplate: 

Here are some sample awards:
https://www.sandbox-freebase.com/m/027y4w0  (an interesting thing about the Nandis is that only a couple categories have gold/silver/bronze: http://en.wikipedia.org/wiki/Nandi_Awards#Gold)

Oops! Forgot Thad's list-of-ranked-winners. More soon.

Jeff Prucher

unread,
Oct 28, 2013, 8:00:07 PM10/28/13
to freebase...@googlegroups.com
Here's a quick-and-dirty ranked winners CVT:
Note that I attached "winner" and "work" to /award/nominee, but that's not really right (not all awards have nominees). Probably we'd need two more types.

Here's an award category with ranked results (note, for the record, that the existing nomination data is incorrect -- this award doesn't have nominees):

I don't know how common this actually is: The Locus and a few other SF magazine-based polls are the only awards I know of where the ranked results are routinely listed.

Thad Guidry

unread,
Oct 28, 2013, 10:10:22 PM10/28/13
to Freebase Discuss
On Mon, Oct 28, 2013 at 7:00 PM, Jeff Prucher <jpru...@google.com> wrote:
Here's a quick-and-dirty ranked winners CVT:
Note that I attached "winner" and "work" to /award/nominee, but that's not really right (not all awards have nominees). Probably we'd need two more types.

Here's an award category with ranked results (note, for the record, that the existing nomination data is incorrect -- this award doesn't have nominees):

The underdogs that I like to see (and that allows my use case... Good Work here Jeff ! )

[{
  "type": "/award/award_ranked_winning_results",
  "rank": null,
  "ns0:rank>": 1,
  "category": "Locus Award for Best Novel",
  "winner": null,
  "work": null,
  "year": null,
  "id": null
}]

 


I don't know how common this actually is: The Locus and a few other SF magazine-based polls are the only awards I know of where the ranked results are routinely listed.


Oh I am sure there are others out there just waiting to be populated.  We have not had this capability, so it will need to be filled out.  But I think it will be immensely useful down the line, perhaps Jon could find some others ?

Now...

How do we cover the use case of saying that "Silver" = Second Place = 2 ?  an alias for it...grrr, not hard lined enough for me to query accurately.  How do we make saying "Silver Clio" is a 2nd place artifact somehow ?  Do we add one more property to the Achievement Level that holds an integer value and then what would we call it ?  "Place of Level" ?  "Position of Level" ?


So the above perhaps could also have a property under Achievement Level called:

"Position of Level" that expects an integer value to note that the "Silver Clio" has a position of 2 for the award "Clio Awards" ?  (along with us also saying that now that it is higher than and lower than...but we still need a concrete position for querying those "2nd placers".

And that "Testimonial" example of the Royal Astronomy would perhaps have a "Position of Level" ...what ?  holding a value of 0 or null I guess ?

Jeff Prucher

unread,
Oct 29, 2013, 6:41:18 PM10/29/13
to freebase...@googlegroups.com
On Mon, Oct 28, 2013 at 7:10 PM, Thad Guidry <thadg...@gmail.com> wrote:

On Mon, Oct 28, 2013 at 7:00 PM, Jeff Prucher <jpru...@google.com> wrote:
Here's a quick-and-dirty ranked winners CVT:
Note that I attached "winner" and "work" to /award/nominee, but that's not really right (not all awards have nominees). Probably we'd need two more types.

Here's an award category with ranked results (note, for the record, that the existing nomination data is incorrect -- this award doesn't have nominees):

The underdogs that I like to see (and that allows my use case... Good Work here Jeff ! )

[{
  "type": "/award/award_ranked_winning_results",
  "rank": null,
  "ns0:rank>": 1,
  "category": "Locus Award for Best Novel",
  "winner": null,
  "work": null,
  "year": null,
  "id": null

}]
 


I don't know how common this actually is: The Locus and a few other SF magazine-based polls are the only awards I know of where the ranked results are routinely listed.


Oh I am sure there are others out there just waiting to be populated.  We have not had this capability, so it will need to be filled out.  But I think it will be immensely useful down the line, perhaps Jon could find some others ?

Now...

How do we cover the use case of saying that "Silver" = Second Place = 2 ?  an alias for it...grrr, not hard lined enough for me to query accurately.  How do we make saying "Silver Clio" is a 2nd place artifact somehow ?  Do we add one more property to the Achievement Level that holds an integer value and then what would we call it ?  "Place of Level" ?  "Position of Level" ?


So the above perhaps could also have a property under Achievement Level called:

"Position of Level" that expects an integer value to note that the "Silver Clio" has a position of 2 for the award "Clio Awards" ?  (along with us also saying that now that it is higher than and lower than...but we still need a concrete position for querying those "2nd placers".

Note that "silver" is not exactly the same thing as second place. See my previous comment about the Clios -- in 2013, for film advertisements, the silver winners are arguably 15th place. I think it's really a parallel model to the whole first/second/third/etc. thing.
 

And that "Testimonial" example of the Royal Astronomy would perhaps have a "Position of Level" ...what ?  holding a value of 0 or null I guess ?

That testimonial example is mighty strange, all right. I'd say it exists in parallel to the silver medal and gold medal, rather than in a hierarchy with them.

Jeff

Thad Guidry

unread,
Oct 29, 2013, 9:51:39 PM10/29/13
to Freebase Discuss
Now...

How do we cover the use case of saying that "Silver" = Second Place = 2 ?  an alias for it...grrr, not hard lined enough for me to query accurately.  How do we make saying "Silver Clio" is a 2nd place artifact somehow ?  Do we add one more property to the Achievement Level that holds an integer value and then what would we call it ?  "Place of Level" ?  "Position of Level" ?


So the above perhaps could also have a property under Achievement Level called:

"Position of Level" that expects an integer value to note that the "Silver Clio" has a position of 2 for the award "Clio Awards" ?  (along with us also saying that now that it is higher than and lower than...but we still need a concrete position for querying those "2nd placers".

Note that "silver" is not exactly the same thing as second place. See my previous comment about the Clios -- in 2013, for film advertisements, the silver winners are arguably 15th place. I think it's really a parallel model to the whole first/second/third/etc. thing.
 

You just saying that all the Golds (14 of them ) and the Grand (1 of them) are all above the Silver ...hence all the Silvers are 15th place ?

That's a strange way to look at them.  I still really do not understand...and I am looking at this page directly: http://www.clioawards.com/catalog/2013/film/

What "position" would you put "Best Job" by Proctor & Gamble for Television/Cinema category ?  It has a Silver... long form... a "Silver Clio" as you state...and that is not the same idea as 2nd place ?  Then elaborate why it would be 15th place.

 

And that "Testimonial" example of the Royal Astronomy would perhaps have a "Position of Level" ...what ?  holding a value of 0 or null I guess ?

That testimonial example is mighty strange, all right. I'd say it exists in parallel to the silver medal and gold medal, rather than in a hierarchy with them.

Jeff

Do we have a model or example for doing these kinds of strange ones in parallel in Schema Staging or Sandbox ?
 

Jeff Prucher

unread,
Oct 30, 2013, 3:07:37 PM10/30/13
to freebase...@googlegroups.com
On Tue, Oct 29, 2013 at 6:51 PM, Thad Guidry <thadg...@gmail.com> wrote:

Now...

How do we cover the use case of saying that "Silver" = Second Place = 2 ?  an alias for it...grrr, not hard lined enough for me to query accurately.  How do we make saying "Silver Clio" is a 2nd place artifact somehow ?  Do we add one more property to the Achievement Level that holds an integer value and then what would we call it ?  "Place of Level" ?  "Position of Level" ?


So the above perhaps could also have a property under Achievement Level called:

"Position of Level" that expects an integer value to note that the "Silver Clio" has a position of 2 for the award "Clio Awards" ?  (along with us also saying that now that it is higher than and lower than...but we still need a concrete position for querying those "2nd placers".

Note that "silver" is not exactly the same thing as second place. See my previous comment about the Clios -- in 2013, for film advertisements, the silver winners are arguably 15th place. I think it's really a parallel model to the whole first/second/third/etc. thing.
 

You just saying that all the Golds (14 of them ) and the Grand (1 of them) are all above the Silver ...hence all the Silvers are 15th place ?

That's a strange way to look at them.  I still really do not understand...and I am looking at this page directly: http://www.clioawards.com/catalog/2013/film/

What "position" would you put "Best Job" by Proctor & Gamble for Television/Cinema category ?  It has a Silver... long form... a "Silver Clio" as you state...and that is not the same idea as 2nd place ?  Then elaborate why it would be 15th place.

It's 15th place because there are fourteen entrants ahead of it. In a 14-way tie for first, the next place is 15th. (Think of it this way -- in the Olympics, if there's a tie for first, no silver medal is awarded; rather there are two golds and a bronze.) 

But I don't think these gold/silver/bronze (or medal/honor) levels necessarily map to ordinals. I think they're just different, hierarchical, levels of award. 
 

 

And that "Testimonial" example of the Royal Astronomy would perhaps have a "Position of Level" ...what ?  holding a value of 0 or null I guess ?

That testimonial example is mighty strange, all right. I'd say it exists in parallel to the silver medal and gold medal, rather than in a hierarchy with them.

Jeff

Do we have a model or example for doing these kinds of strange ones in parallel in Schema Staging or Sandbox ?

This is a great use case, Thad!

I'm not entirely sure what the right approach is here. There are three options, none of which seems necessarily better than the others:

1. Treat them as separate awards entirely. This isn't entirely satisfactory since they're clearly related.
2. Treat them as different categories of a "Medal of the Royal Astronomical Society" award.
3. Treat them as different levels of achievement of a dual "Royal Medal of the Astronomical Society" award/award category. (Which looks like https://www.sandbox-freebase.com/m/0y4y64j. Note that gold and silver are above and below each other, but testimonial is not above or below anything.)

#2 hurts my head the least, although we lose the hierarchy.
#3 maintains the hierarchy, but makes it harder to find the medalists (the path from "gold medal of the RAS" is different from the usual category-to-winner path).

So maybe the achievement level isn't the best model, and we should go back to having them all be /award/award_category, with a higher/lower property on the category itself? I'm torn, because I don't know what to do with the Clio nominees; without a separate achievement level property, there's no where to stick them. I guess we could just rule that nominees are always at the top level, and you could walk down the hierarchy to find which nominees actually won something. That might be simplest.

Here's what that might look like (I only did gold and silver to save time): 
silver: https://www.sandbox-freebase.com/m/0y4y6h2 (note nominees are has_no_value)

Jeff
 

--

Thad Guidry

unread,
Oct 30, 2013, 3:37:15 PM10/30/13
to Freebase Discuss
On Wed, Oct 30, 2013 at 2:07 PM, Jeff Prucher <jpru...@google.com> wrote:



On Tue, Oct 29, 2013 at 6:51 PM, Thad Guidry <thadg...@gmail.com> wrote:

Now...

How do we cover the use case of saying that "Silver" = Second Place = 2 ?  an alias for it...grrr, not hard lined enough for me to query accurately.  How do we make saying "Silver Clio" is a 2nd place artifact somehow ?  Do we add one more property to the Achievement Level that holds an integer value and then what would we call it ?  "Place of Level" ?  "Position of Level" ?


So the above perhaps could also have a property under Achievement Level called:

"Position of Level" that expects an integer value to note that the "Silver Clio" has a position of 2 for the award "Clio Awards" ?  (along with us also saying that now that it is higher than and lower than...but we still need a concrete position for querying those "2nd placers".

Note that "silver" is not exactly the same thing as second place. See my previous comment about the Clios -- in 2013, for film advertisements, the silver winners are arguably 15th place. I think it's really a parallel model to the whole first/second/third/etc. thing.
 

You just saying that all the Golds (14 of them ) and the Grand (1 of them) are all above the Silver ...hence all the Silvers are 15th place ?

That's a strange way to look at them.  I still really do not understand...and I am looking at this page directly: http://www.clioawards.com/catalog/2013/film/

What "position" would you put "Best Job" by Proctor & Gamble for Television/Cinema category ?  It has a Silver... long form... a "Silver Clio" as you state...and that is not the same idea as 2nd place ?  Then elaborate why it would be 15th place.

It's 15th place because there are fourteen entrants ahead of it. In a 14-way tie for first, the next place is 15th. (Think of it this way -- in the Olympics, if there's a tie for first, no silver medal is awarded; rather there are two golds and a bronze.)  

But I don't think these gold/silver/bronze (or medal/honor) levels necessarily map to ordinals. I think they're just different, hierarchical, levels of award. 
 

OK that makes sense now.  Thanks for explaining that.
 

 

And that "Testimonial" example of the Royal Astronomy would perhaps have a "Position of Level" ...what ?  holding a value of 0 or null I guess ?

That testimonial example is mighty strange, all right. I'd say it exists in parallel to the silver medal and gold medal, rather than in a hierarchy with them.

Jeff

Do we have a model or example for doing these kinds of strange ones in parallel in Schema Staging or Sandbox ?

This is a great use case, Thad!

I'm not entirely sure what the right approach is here. There are three options, none of which seems necessarily better than the others:

1. Treat them as separate awards entirely. This isn't entirely satisfactory since they're clearly related.
2. Treat them as different categories of a "Medal of the Royal Astronomical Society" award.
3. Treat them as different levels of achievement of a dual "Royal Medal of the Astronomical Society" award/award category. (Which looks like https://www.sandbox-freebase.com/m/0y4y64j. Note that gold and silver are above and below each other, but testimonial is not above or below anything.)

#2 hurts my head the least, although we lose the hierarchy.
#3 maintains the hierarchy, but makes it harder to find the medalists (the path from "gold medal of the RAS" is different from the usual category-to-winner path).

So maybe the achievement level isn't the best model, and we should go back to having them all be /award/award_category, with a higher/lower property on the category itself? I'm torn, because I don't know what to do with the Clio nominees; without a separate achievement level property, there's no where to stick them. I guess we could just rule that nominees are always at the top level, and you could walk down the hierarchy to find which nominees actually won something. That might be simplest.


Instead of walking... Could we also just add a WINNER boolean flag property to the award/award_nomination CVT ?  saves the walking ?  If you look at it...mostly all the properties are under nomination already... so the main data lies there... https://www.sandbox-freebase.com/m/0y4y6fw and then your just flagging which of the nominees WON, in that case, Barton F. Graf.
 
Here's what that might look like (I only did gold and silver to save time): 
silver: https://www.sandbox-freebase.com/m/0y4y6h2 (note nominees are has_no_value) 

And why does nominees has_no_value ?  Because there were no nominees in this case ?

-- 

Jeff Prucher

unread,
Oct 30, 2013, 5:35:46 PM10/30/13
to freebase...@googlegroups.com
The trouble is that not all awards have nominations (Nobels, Newberys, Pritzkers, Gold Medal of the RAS, etc.). To say that the winners of those awards were also nominees would be incorrect. A practical effect of that incorrectness would be that anyone attempting to compare nominees to winners would be unable to (this isn't entirely hypothetical -- I curated a bunch of science fiction and fantasy awards in Freebase a few years ago, partly to try to see if I could compare nominations vs wins by gender; if all winners were also nominees, it would have skewed the results). 

 
Here's what that might look like (I only did gold and silver to save time): 
silver: https://www.sandbox-freebase.com/m/0y4y6h2 (note nominees are has_no_value) 

And why does nominees has_no_value ?  Because there were no nominees in this case ?

Because, in this proposal anyway, I think that the nominations should be all be attached to the top-level category ("Clio for Audio - Gold" in this case), and asserting has_no_value on the silver category is signal that you shouldn't be sticking nominees there. It would also be signal that if someone did, it would need to be cleaned up. (We'd have documentation and stuff, too, of course.)
 
Jeff

--

Thad Guidry

unread,
Oct 30, 2013, 5:45:07 PM10/30/13
to Freebase Discuss
 
Here's what that might look like (I only did gold and silver to save time): 
silver: https://www.sandbox-freebase.com/m/0y4y6h2 (note nominees are has_no_value) 

And why does nominees has_no_value ?  Because there were no nominees in this case ?

Because, in this proposal anyway, I think that the nominations should be all be attached to the top-level category ("Clio for Audio - Gold" in this case), and asserting has_no_value on the silver category is signal that you shouldn't be sticking nominees there. It would also be signal that if someone did, it would need to be cleaned up. (We'd have documentation and stuff, too, of course.)
 
Jeff

OK, I am in agreeement there then... having a has_no_value as the signal.  And documentation is always good.

Jeff Prucher

unread,
Oct 31, 2013, 4:46:06 PM10/31/13
to freebase...@googlegroups.com
I've been talking off-list with the rest of the schema team, and there is universal agreement that #2 above is the best approach, since it preserves the semantics of what an award category is, and especially because of being able to link nominations correctly (nobody bought my "link it to the top level" story). :( Which is to say the schema originally proposal upthread:

> I made a schema on sandbox, since trying to extend a CVT on 
> /base/schemastaging was going to be too hideous to contemplate: 
>
> Here are some sample awards:
> https://www.sandbox-freebase.com/m/027y4w0  (an interesting thing about the 
> Nandis is that only a couple categories have gold/silver/bronze: 

So I'm going to stick this on schemastaging as best I can. I'm happy to keep working through use cases, but since we need to add a new property to a CVT, I'd like to put it in /award sooner rather than later (since Jon's team is working on awards at this very moment, and dealing with extended CVTs in schemastaging is a big fat PITA).

Jeff
 


Instead of walking... Could we also just add a WINNER boolean flag property to the award/award_nomination CVT ?  saves the walking ?  If you look at it...mostly all the properties are under nomination already... so the main data lies there... https://www.sandbox-freebase.com/m/0y4y6fw and then your just flagging which of the nominees WON, in that case, Barton F. Graf.
 
Here's what that might look like (I only did gold and silver to save time): 
silver: https://www.sandbox-freebase.com/m/0y4y6h2 (note nominees are has_no_value) 

And why does nominees has_no_value ?  Because there were no nominees in this case ?

-- 

--

Thad Guidry

unread,
Oct 31, 2013, 5:05:24 PM10/31/13
to Freebase Discuss
Go forward then with those plans, Jeff.  I think there is sense in it and agree to add the new property to the CVT sooner (now).

(And I am also glad to have helped in discovering some of the use cases ! )

[snip]



I've been talking off-list with the rest of the schema team, and there is universal agreement that #2 above is the best approach, since it preserves the semantics of what an award category is, and especially because of being able to link nominations correctly (nobody bought my "link it to the top level" story). :( Which is to say the schema originally proposal upthread:

> I made a schema on sandbox, since trying to extend a CVT on 
> /base/schemastaging was going to be too hideous to contemplate: 
>
> Here are some sample awards:
> https://www.sandbox-freebase.com/m/027y4w0  (an interesting thing about the 
> Nandis is that only a couple categories have gold/silver/bronze: 

So I'm going to stick this on schemastaging as best I can. I'm happy to keep working through use cases, but since we need to add a new property to a CVT, I'd like to put it in /award sooner rather than later (since Jon's team is working on awards at this very moment, and dealing with extended CVTs in schemastaging is a big fat PITA).

Jeff
 
 
Reply all
Reply to author
Forward
0 new messages