Freebase: Discussion about spatialed

Biology: Acre applications to populate data?

2010-03-18T22:49:50.0000Z

There are additional issues with this that need to be fixed before it's usable, so I've disabled the species-to-genus option until someone has time to repair it.

Any genus without its scientific name property filled in won't get resolved (pretty common for a genus which has the same common name as scientific name).
Any genus that gets created doesn't get used as part of the resolution process.

The second problem could be fixed just by running the tool a second time, but the first problem is causing lots of duplicate genii to be created. It shouldn't be used for species until this is fixed (and all the duplicates that have already been created need to be flagged for merger).

If I get a few minutes I'm going to focus on identifying and flagging the things that need to be merged as the first priority.

Biology: Acre applications to populate data?

2010-03-18T22:08:16.0013Z

I made the one line fix, but I'm having problems testing it because it continually times out. The one time I did get it to complete it was working on Genus that existed so it didn't test the code path I wanted. If anyone else tries it and finds problems, let me know.

I snapshotted version 2 before my changes, so you can fall back to that if there are problems. http://2.commonsense.doconnor.user.dev.freebaseapps.com/

Are people happy with the methodology of assigning species to a genus which is the first word of the species scientific name? If so, it'd be easy enough to code this up in Python so that it doesn't run into the timeout problems.

Biology: Acre applications to populate data?

2010-03-18T07:20:26.0024Z

Oh oh.

I've got a few (non freebase) things to knock off before I can get to this meaningfully - if you wish, I've added you with access to the application; please feel free to make any changes you deem necessary.

Biology: Acre applications to populate data?

2010-03-18T06:58:07.0047Z

The commonsense tool needs to be changed to add the type /common/topic on newly created topics. Without this they don't show up in autocomplete and other basic operations like merge don't work.

I just submitted fixes for a batch of almost 6K of these which should clear up the current stuff, but the tool should be fixed if it's going to be used on an ongoing basis.

Biology: Acre applications to populate data?

2010-03-18T00:37:02.0013Z

about 2% of the 86k names I asserted didnt match ^[A-Z][a-z]{2,} [a-z]{3,}$' so ill add that check.

the single letter culprits are: 90 U. 23 P. 19 C. 6 M. 6 D. 6 B. 5 N. 5 A. 3 S. 2 T. 2 L. 2 G. 1 H. 1 E.

Biology: Acre applications to populate data?

2010-03-18T00:23:56.0013Z

good catch, sorry about that.

Biology: Acre applications to populate data?

2010-03-17T22:49:39.0013Z

Also; there's a fair few names like "C. (something)" which should be cleaned up - doing by hand now; but if you wanted to build this into the next iteration of the tool...

[{ "FBID430:scientific_name": [{ "lang": "/lang/en", "type": "/type/text", "value~=": "^C\.*" }], "id": null, "name": null, "scientific_name": null, "type": "/biology/organism_classification" }]

Biology: Acre applications to populate data?

2010-03-17T22:29:18.0013Z

Oops! We got a few extra bits like wiki markup for italic text or links.

I'm cleaning up by hand.

// MQL to see leading quotes [{ "FBID430:scientific_name": [{ "lang": "/lang/en", "type": "/type/text", "value~=": "^\'*" }], "id": null, "name": null, "scientific_name": null, "type": "/biology/organism_classification" }]

Company: Good Grief - Business Locations property

2010-03-16T02:23:29.0006Z

Thanks Jeff. I look forward to better "tools" to upload and work with data.

Company: Good Grief - Business Locations property

2010-03-15T23:00:44.0013Z

I think that "Y Store, Fifth Ave., New York" is pretty acceptable, assuming that there are other NY locations. (Doing it for businesses with a single location would be unnecessary, I should think.) It's pretty common, I think, to refer to individual stores (restaurants, branch offices, etc.) this way (the "New York office", the "Union Square store", etc.). The alternative would be to have hundreds of topics named "Starbucks" with no clear way of disambiguating them, since the only real disambiguating property is the address, and since address is a CVT, it doesn't show up like other disambiguating properties (e.g., in the autocomplete flyout) . Having the full address in the name like happened with the ChefMoz import is probably overkill, though.
The properties that are available when you're creating a new topic from that particular view are those which are displayed in the table. These automatically-generated views display some combination of the first few properties of the type (ordered as they are when you look at the schema or at a topic in edit mode) and the properties marked as disambiguators (I don't know the exact logic). Note also that properties that expect CVTs are not included when determining which properties to offer for input on this page. So, of the properties on Business Location, I think the ones being presented are pretty sensible. I'm not a UI designer by a long shot, but as long as you can only create one new topic at a time, it doesn't seem any harder to me to enter the properties vertically than horizontally -- you still have to tab (or click) between them, and at least with vertical input, you don't get the side-scrolling that occurs for some properties when you edit them from the topic page.

But requests for UI improvements can be made in JIRA; like I said, I'm not a designer, so I only have limited insight into these issues. I did notice, however, that the property hints aren't displayed (even as tooltips) when creating topics from the table view page, which is clearly an oversight, so I opened a ticket for that: CLI-9745.

Company: Good Grief - Business Locations property

2010-03-15T21:15:29.0013Z

I might get a bit off topic here concerning the Web UI itself, but bear with me...

2 problems were identified by me, Jeff.

Should the tool tip description remain as giving an example of "Y Store, Fifth Ave., New York" that is perfectly acceptable , when we clearly don't want redundancy, but merely normalization ? ....and if so # 2 is a suggestion to improve the data collection UI ...
Why couldn't the Web UI be improved ? When I go here to this view to add business locations http://www.freebase.com/view/business/business_location and click add more topics, the pop-down defaults to showing only the name, opening date, closing date, and parent company. I think better properties to collect here would be the location details themselves, yes ?, such as name, address, city, state, and parent company and preferrably show them going left to right as a new row, just like a "gasp" spreadsheet, rather than vertically now on the right? The UI is really the culprit for me, when I'm going to add less than 10 items, because it doesn't mimic a spreadsheet view entirely, in my opinion. For more than 10 items, I use an external spreadsheet editor and tools such as Loader.

I stay in EDIT mode, so I benefit, but we cursed sometimes by that view and it's tempting "edit" link on the left of individual properties on any particular topic and usually smack face first into data collection problems from a users perspective such as this. I rarely click the Detailed View link just to the right of it. Don't get me wrong, I like the inline edit, but sometimes, its ... a pain, and slow.

Company: Good Grief - Business Locations property

2010-03-15T20:10:41.0015Z

This is a complex thread to jump in late to, but I think we're down to just the property name being an issue? (Since the help text wasn't causing the topic names which included street addresses, which I think was the main cause of Thad's complaint about them. But I could be wrong, and if there are additional ways the help text could be improved, though, I'd be happy to hear them.)

The problem with the property name, as Tom says, is that the Business Location type can be used for all kinds of things, so we need a term that can encompass branches, headquarters, outlets, store/restaurant locations, offices, and the like. Unfortunately, the closest term, "location," has another meaning which is confusing. (Certainly, in the retail world, individual stores of a chain are commonly called "locations.") We could go in the direction of a list, and call it something like "Branches, stores, etc." I try to avoid that sort of phrasing in general, but sometimes it's the best choice. Other suggestions would be welcome, too.

Time: Make dates unique

2010-03-15T19:44:58.0006Z

I spent a bunch of time poking around Wikipedia, trying to see if there was an easy way to extract the dates from the infoboxes, but the date range information is tucked away somewhere not easily extractable, alas.

Time: Make dates unique

2010-03-14T20:29:46.0015Z

Thanks for fixing this up. I typed a bunch more topics with the newly revamped type, but many of them still need their dates filled in.

Company: Good Grief - Business Locations property

2010-03-14T17:09:23.0013Z

You can name your new Business Location "Loving Care," "Loving Care branch," or "Loving Care, Jersey Shore." It's just a matter of taste, not something critical. Personally, I'd pick something that made sense in the local context as opposed to something that helps with global disambiguation.

The Burger Kings were imported from ChefMoz and someone made the decision to tack the addresses onto the name. It's not something that was done manually and it'd be easy enough to script a fix to remove them if it was important enough to someone. On ChefMoz itself, they're listed with just the name "Burger King" which is what would make sense to me for Freebase. Note that they all have address information encoded in properties too, so the information in the name is redundant.

The problem with choosing a new name for the property is finding one which will work in all the contexts where this can be used. It's wider in scope than just "branch" or something equally simple. I'm afraid I don't have any good suggestions for alternatives.

Company: Good Grief - Business Locations property

2010-03-14T16:41:27.0000Z

I agree the name should change. Not sure to what, but lots of folks get confused by it

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-13T19:12:21.0013Z

Actually - after playing around with: http://www.itis.gov/ITISWebService/services/ITISService/getRecordFromLSID?lsid=urn:lsid:itis.gov:itis_tsn:720256

I appears that the services returns XML if your accept header specifies XML (any string with XML in it?)

I'm thinking there should be a CVT on the Consuming API that expresses what accept headers produce what types of output.

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-13T18:53:05.0006Z

Oh for dumb (me!) I looked at the output, thinking it was a redirect in my browser - but failed to do the obvious check with RDF headers.

By simply changing the "Consuming API Format" it will prevent itis.gov link showing up in the RDF output, but still operate as a way of documenting the /biology/itis namespace as linking these two data sets (if even in a non-machine readable way.)

It is interesting to note that the sameas.freebase.com base was designed to support RDF sameAs generation, but not to be used exclusively for it. The hope was it would become a directory for any API where you could construct calls and identifiers from Freebase keys.

The iphylo link is interesting, I'd love to link it in! How does the NCBI Taxon ID relate to ITIS? (When I request http://bioguid.info/taxonomy/720256 as RDF I get "Lepidoptera sp. BOLD:BIN60780" which seems to be different than the Agelastica alni beetle...)

PS - I started work on the sameAs URI Template application in ACRE - thanks for the idea, I'll see if I can get a version out next week.

Company: Good Grief - Business Locations property

2010-03-13T17:24:08.0000Z

Yes, Tom, that was the tool tip I looked at. I don't like the fact that I have to click on the ?. But besides that, I still don't understand the helpfulness of this property if it doesn't explain itself well enough. That last sentence is the key to good data, in my opinion. We have seen what happens when a property is NOT described at all, even you and I have been bit by that scenario. Certainly, I am not going to fill it with all the McDonalds, Burger King, or Wal-Mart locations. (Someone appears to already be doing that - lol.)

If I understand this correctly, your saying that for Loving Care Agency and it's 28 business locations( but I'm NOT supposed to think of it as a LOCATION, even though it uses the term, I'm supposed to think of this as a branch office - lol), that I would type in a unique name, such as "Loving Care Office, Jersey Shore" and it shows as that business location type is not found, of course, so I then click CREATE NEW BUSINESS LOCATION, which creates a new unique GUID and TOPIC for a business_location , and then for that new business_location store, office, branch, etc that was created, at it's /business/business_location/address I would add the address information filling in the appropriate city there. I'm fine with that.

However, The bad tool tip description, to me, results in extra information that is sometimes not needed. In my opinion, storing address information freeform in the name misses the whole point of Freebase. And results in illogical views such as this http://www.freebase.com/view/en/burger_king/-/business/company/locations where you have the address field already showing and certainly you can create your own views whenever you want.

I think reducing the ambiguous description will also help reduce these extremely long business location names that really aren't necessary. THIS BAD DESCRIPTION IS CREATING VERY LAZY FOLKS that type a bunch of properties, address, city, state, etc. all in one line, the "business location name", instead of taking the time to fill in the appropriate properties, and collect good data. Or perhaps they didn't properly scrub and delimit their data loads, who knows. But in my opinion, 1 good bit is worth more than 1000 bad extra bits.

What are your ideas or agreements ?

Company: Good Grief - Business Locations property

2010-03-13T00:06:25.0013Z

I don't know if it's been edited since you posted this, but the help for the property currently says "Offices, stores, outlets, plants, campuses, etc. run by this company. These might not always have obvious names, but names like "X Company Headquarters" or "Y Store, Fifth Ave., New York" are perfectly acceptable. (Enter Business Location)". This is accessible by clicking the ? in the data entry popup. Is that what you were looking at or some other tool tip?

Think of it as a branch office, not a location like a city or street.

Company: Good Grief - Business Locations property

2010-03-12T22:15:04.0011Z

I really thought I was much smarter than I appeared to be... until today.

My assumption was that Business Locations could include City Names. Apparently, I was wrong. See Loving Care Agency

I made this mistake, because the tip text mentioned "NY", but apparently, the implied meaning is that this field should be filled with a specific business location NAME. Such as "Perth Amboy, Adult Care Center, NJ" and not just type and link directly to a /city topic such as Perth Amboy

I am wondering how many others get confused by this property ? And if the help text should be changed ? I don't like the way it links back and messes with the wrong type of view when looking at City's. Maybe just the linking needs to be changed somehow to make this look better ? Perhaps the Contains property instead ? For example:

[Perth Amboy] (http://www.freebase.com/view/en/perth_amboy) could have a nice view showing which companies have a business location there? Was that they intent here ? or should we just drop this property altogether, and instead use "Contains" property to favor better use of Lat/Long/Geocodes ?

Anyone have thoughts on a better way to handle "Business Locations" ? Or, perhaps it just needs to be expanded with a CVT and the city ?

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-12T08:11:00.0006Z

re change = new LSID...

What's the main rationale behind that?

Scenarios like "we thought it was a fungi, oops, it turned out to be a new kind of bacteria"?

Perhaps for freebase's use case where it's mapping a wikipedia article to an ITIS entry, if the ITIS entry changes / their LSID changes, we'd (hopefully) pick that up via automated processes and update to the new ITIS entry?

Biology: Updated type definition for Organism Classification

2010-03-12T06:14:02.0013Z

No, not generally.

It's just that in this case, a Coydog is not really a Breed at all -- as far as I can tell, they are not generally raised or bred. They are a hybrid of Coyotes and Dogs.

Biology: Acre applications to populate data?

2010-03-11T23:45:08.0013Z

I think you are dead right in that we need to provide more value. At the moment freebase offers a pre-parsed version of wikipedia content (warts and all), mapped to some existing data sources (ie: ITIS), and a query language with a low barrier to learn (mql, not unlike sparql, basically json); as well as RDF output linked back to dbpedia/other places (linked data yay).

The next level of offerings that we need to get to are:

A bunch of really slick tools which can assimilate other unique large data sources for low effort (ie: http://data.australia.gov.au/570)
Interlink relevant data extremely densely, focusing on identifiers (so we can help power http://sameas.org/) - this is already done by some
Simple but effective visualisation tools which are more focused on the general public / data meshups which are relevant and interesting to the general public - ie; we know about human populated urban areas (things typed with /city/town); can we provide the 'top 5 most common' animals in a given area (and push people off to learn more on other sites)? Can we, for instance, build decent indexes of animals zoos of the world; so that we can answer the question of "where can I see this animal"?

After we have that kind of thing in place, if there's no real interest, I'll personally kidnap a few people in the field; and go back to the "lock them into the room" plan :D

Biology: Acre applications to populate data?

2010-03-11T22:49:20.0013Z

I added 48k scientific_names using doconnor's suggestion.

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-11T20:32:45.0018Z

People may want to review what LSIDs are. They are not like standard identifiers. For instance if there is any change in the entity described by a LSID it gets a new LSID.

I don't think they make stable identifiers for the semantic web.

Pete

Biology: Updated type definition for Organism Classification

2010-03-11T00:36:05.0013Z

So as a general rule, Hybrid should take precedence over Breed?

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-11T00:30:06.0013Z

Sorry, my mistake - misread the original question :(

Biology: Updated type definition for Organism Classification

2010-03-11T00:26:59.0000Z

Yes -- you're right. In fact, it's more a Hybrid than a Breed. Made the change.

Biology: Updated type definition for Organism Classification

2010-03-11T00:14:34.0000Z

Isn't a Coydog also a hybrid?

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-10T22:43:55.0000Z

Woo! I can't decide if it should be "fb.common_topic owl:sameAs urn:lsid:itis.gov:itis_tsn:720256" ; which is bad for linked data people in that you can't fetch it (but others can state that something is rdf:about that urn or owl:sameAs it);

or "fb.common_topic owl:sameAs (resolver link)", like you've done - which unfortunately returns a non RDF resource (and I don't think there's anything you can do to GRDDL it.

But at least there's something to start off with :)

Biology: Updated type definition for Organism Classification

2010-03-10T20:14:13.0000Z

Hi pak21 --

Looks like the evles waved their magic wand and made me an admin of the Biology domain ... ;-)

Breed has been added as an Organism Classification Rank, with higher rank Species and Subspecies. Labs and Coydogs have been marked as Breeds.

This opens up one final question -- should /biology/domesticated_animal/breed delegate from /biology/organism_classification_rank? Or do we need to set up a gardening task for this?

Brian

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-10T19:07:28.0005Z

Daniel -

You are right - creating a Consuming API and URI Template for other services is very cumbersome. I will take you up on your wish and see if I can cobble something together in ACRE that makes generating these entries easy.

In the meantime, I've added an ITIS.gov consuming API entry and a corresponding URI Template to generate the sameAs link in the Freebase RDF.

Now if you request the RDF for /biology/itis/720256

with someething like: wget -O - --header="accept:application/rdf+xml" http://rdf.freebase.com/ns/biology.itis.720256

You should find an itis.gov owl:sameAs triple in the output:

jeff: Director property

2010-03-10T18:11:22.0013Z

Yeah, I'm not crazy about the writer denormalization, either, since all we're doing is saying "people who have done X thing more than (some threshold)." (The producer and actor ones are a bit different, since the role of the producer of an entire program vs. a single episode is different, as are the semantics of a "starring" acting role vs. appearing in an episode.) Let's run it by the discuss list and see if we can get more input there.

Biology: Updated type definition for Organism Classification

2010-03-10T17:36:56.0000Z

Hi pak21 --

The first point easily fixed in the data ... in fact, it looks like some elf somewhere already took care of Kokoro and Pyamidata.

All the rest of the points seems to revolve around the proper coding of Breed as an Organism Classification Rank. I note we already have Subspecies, Varietal, Cultivar, Morph and "Discrete Population Segment", so it seems like we should have Breed.

I'm not an admin in the domain, so I can't add it to the Organism Classification Rank enumeration. I assume you are an admin of this domain -- would you do the honors?

Brian

Location: event location

2010-03-10T16:56:49.0000Z

thanks!

Biology: Updated type definition for Organism Classification

2010-03-10T11:19:41.0006Z

That's a lot better, and certainly better than the copyright violation :-) The main problem I have with it now is that the examples given disagree with the data on Freebase:

Kokoro and Pyramidata aren't typed as organism classifications at all.
Labrador Retriever and Coydog are typed as a organism classifications, but don't have a higher classification linked
The text is internally inconsistent: the table claims that a Labrador Retriever is a breed of the species Canis lupus, which what the data in Freebase claims, but the text later on says "all domesticated dogs are grouped in the same species, Canis familiaris". Which one is it? (Yes, I know this is a tricky question, but at least let's not try to confuse people even more!)
Is Breed an organism classification rank or not? The text very much claims that it is, but it's not actually typed that way on Freebase.

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-10T08:17:13.0013Z

Luke: currently there's nothing in that RDF which explicitly says that what Freebase knows as /biology/itis/720256 is actually the same thing as http://www.itis.gov/ITISWebService/services/ITISService/getRecordFromLSID?lsid=urn:lsid:itis.gov:itis_tsn:720256 ; the idea here is to add that link.

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-10T07:15:14.0000Z

The documentation around creating URI templates and so on and so forth is less than clear :(

wishes for an acre application to do it for him in a guided manner

jeff: Director property

2010-03-10T05:50:43.0013Z

The problem is that I don't have episode data, just the tv program level data. As we already have this denormalisation for actors, writers and producers, it would be useful to have it for directors as well.

Location: Partially overlapping locations

2010-03-10T04:58:05.0000Z

Yeah, just after posting this I discovered Geosearch. I'm very impressed.

I think Ed's idea is a good one for its simplicity. My idea was more towards batch-populating this new property from information that can be derived from existing properties (i.e. all-pairs intersection query on the GIS database). Perhaps a decent opportunity for an app?

jeff: Director property

2010-03-09T23:50:31.0013Z

What about using eMQL for this? Since the data is already there (that is, on the episodes themselves), using eMQL to collect them and expose them at the program level might work, and spare us the semi-denormalization. (This is a suggestion, mind you; I haven't really played around with eMQL much.)

Biology: LSID, owl:sameAs / rdfs:seeAlso for ITIS IDs on organism classifications?

2010-03-09T23:14:34.0013Z

This may or may not be relevant, but there's an RDF button on the bottom of all Freebase topics (in edit mode :) which takes you to a list of RDFs for that topic, e.g. http://rdf.freebase.com/rdf/biology.itis.720256.

Biology: Updated type definition for Organism Classification

2010-03-09T23:03:44.0013Z

Hi pak21 --

All of your concerns are valid. We've tried to incorporate them into the wiki:

http://wiki.freebase.com/wiki/Commons/biology/organism_classification

Could you give it a read and let us know what you think?

Brian

Location: event location

2010-03-09T22:54:16.0005Z

Yes, good idea. This is something that had been kicked around so many times, we sort of lost track of it. I've added the reverse property to Location now.

Location: event location

2010-03-09T19:20:42.0000Z

jeff?

jeff: Uniquify Founded

2010-03-09T18:16:57.0013Z

Good catch. Added to the current uniquification task, DA-976.

Location: Partially overlapping locations

2010-03-09T17:29:00.0006Z

For places that have shapes specified, the Geosearch API should allow you to find intersecting locations, but I think Ed was asking for something simpler.

jeff: Uniquify Founded

2010-03-09T17:17:26.0000Z

The /broadcast/radio_station/founded property should be uniquified.

Organism Classification Rank: Higher/lower rank

2010-03-09T12:53:36.0030Z

http://en.wikipedia.org/wiki/Taxonomic_rank

Tada; heirachy explained!

Location: Partially overlapping locations

2010-03-09T12:50:07.0030Z

This would be a good idea; especially if it could be presented as a synthetic property derived from geolocation boundary intersections where present/possible.

At the same time, I realize this would qualify as a very intensive GIS application :)

Location: event location

2010-03-09T12:47:42.0032Z

I second this idea.

jeff: Director property

2010-03-09T12:20:27.0034Z

bumping. No action on this, and I still require this property.

Biology: Updated type definition for Organism Classification

2010-03-09T09:41:52.0038Z

Anne,

It would be appreciated if you could reply to the concerns I and others have raised above.

Thanks.