Discussions on Government

OK, will do.

Out of curiosity, how did you find that out?

Hi tim_mcnamara.

If you take a look at the schema of Government (far right tab next to News and Members), you'll see some Types like Governmental Body and Government Agency.

It looks like there is a type for Legal System, which would be for a judicial system. Otherwise, there's a Type for specific courts. These are not in the Government domain, but instead under Law.

As for despotism, it is considered a Form of Government and is typed as such.

I hope this helps. If you have any other questions, don't hesitate to ask.

Just to clarify. Are you talking about creating additional Freebase types or additional "types of government"? There is already a Freebase type (Form of Government) which has over 50 "types of government". Every country topic in Freebase is able to specify one of these forms of government.

In theory, each of these could be it's own Freebase type but in general we don't like to add new types unless there is significant data which can't be modeled using the existing types. What would you like to model using this data?

For reference, I've created a topic "Australian Senate election in Tasmania, 2007" (following the Wikipedia convention - see e.g. United States Senate election in Arizona, 2000) and filled in the senators elected and some other details.

Good to know that people are thinking about Australia! :)

Revisiting this issue for European Parliament elections: how should we model the results of the elections at a country-by-country level? The Wikipedia page on "European Parliament constituency" states that "In six EU member states (Belgium, France, Ireland, Italy, Poland and the United Kingdom), the national territory is divided into a number of constituencies. In the remaining member states, the whole country forms a single constituency."

So should the European Parliament election, 2004 for Portugal (e.g.) be an Election Contest, but the European Parliament election, 2004 for Ireland be a General Election?

Works for me.

A hint for getting help: explain the actual problem you're trying to solve, rather than what you think the solution is. At the moment, it's rather difficult to help you because I'm guessing what you're actually trying to do.

Or this: http://tinyurl.com/28m85f4.

Reassuringly, they both give the same answer. It's an open question whether a) the full political history of all US presidents has been entered or b) the various state legislature topics have been correctly linked to their jurisdictions and "government office or title" topics, however.

No. Lincoln was a state lower house legistlatur not a state senator.

If changed "name~=": "Senate" to "name~=": "House" I will get the same results except Theodore Roosevelt because New York Legislature (en/new_york_legislature) consists of "New York State Assembly" and "New York State Senate" ( i.e. Assembly instead of House).

Looking at a Google-translated web page, it does seem to be an indirect election.

The big difference for me between indirect and direct elections is who does the voting: a direct election implies universal suffrage of the "citizens" (however that is defined) of the jurisdiction in question. So I would prefer to have a new type for indirect elections.

@Luke: I agree that that's the most important distinction. But if we create a parallel model for indirect elections, we start to run into something that's going to look like a denormalization*. Districts will need two properties for elections: one for direct and one for indirect; politicians will have two incoming properties for their elections. Anyone wanting to find out what election(s) someone was elected in or the outcomes of elections from a district would have to make two queries. Is there a way to represent this by modifying the current schema? The only two things I can think of off-hand (which aren't really that good, IMO) are a boolean or enumeration to hold the Direct or Indirect value, and an Indirect Election co-type that would be added to Election Contest or General Election. The former would be rather annoying to get data into, since every single election would need to have it filled in; the latter only really makes sense if there are properties specific to indirect elections, but I don't know of any. Other thoughts?

*It wouldn't technically be, because the semantics of the two types are different, but to a naive user or process, it's going to look like one.

Good point - you're right, it would be better not to have a separate type.

Here's one idea: add a property to the General Election (e.g. type) called something like Electors; for indirect elections, this would be the actual group that votes (e.g. for French Senate elections, it seems to be Grands électeurs français - can't find the corresponding page in the English Wikipedia, while for elections of the Chief Executive of Hong Kong, it would be the Election Committee), but for direct elections, it would be assumed to be the voters in the political district (so could be explicitly given as such if desired).

Otherwise the electors could be a property just on indirect elections.

What do you think?

Sorry... Perhaps I should really just say the kinds of things that I foresee Freebase (or a database in general) help to answer questions about elections such as :

Did the Multiparty system options perform better or worst in higher literacy levels or areas within Nepals 75 districts ?

Wikipedia article states..." In general, the multiparty system option performed better in areas with higher literacy levels. "

Being able to collect and store in Freebase those percentages FOR and AGAINST in the 75 districts of Nepal. And thus answering the above question.

Hi Dublinstreams,

To add politicians to Freebase, go to Politician and click on the "Add More Topics" button (upper right). Enter each name, or else click on "Go to the import tool" to paste in a list. Once you've added them, you can edit each one to specify which governmental body they're part of. You should do this using the "Government positions held" property; the 30th Dail would be under "Legislative Sessions".

You could also start from 30th Dail and edit "Members" there. If you wanted to import a list, you could click on the little arrow next to the "Edit" button and choose "Import list".

okay, might do a bit more fiddling in sandbox to learn instead of here, now that i found it, you don't make it easy to find your help content... or basic start advice.

so will you pull in their constituency and positions and date of assumed office too? dob etc etc

I was wonder having connected to the wiki, why you didn't pull in the other info too...

the problem is that the infoboxes on wikipedia most poltician pages, they have "assumed office" for both the date they win an election and they date they are appointed to particular position in government or party, which is quite unclear... i was hoping to scrap the info for timeline, but even dbpedia uses termstart for both coming from wikipedia... as will you's i suspect...

have to say how long should i wait, often website promise something soon as years pass by

You're preaching to the choir as far as the "Getting started" docs go. Another good source of information which you may not be aware of is the archives for the 'developers' and 'data-modeling' lists. The wiki is new, so it's not well populated yet. The good news is - it's a wiki, so as you find stuff that you think should have been there, go ahead and add it. It's hard to replicate the viewpoint of a fresh set of eyes.

As far as timing goes, the developer who's working on the politicians already did a trial load on the sandbox (although it didn't include the TDs because the topic wasn't typed correctly yet), which usually means things are pretty close. I'd guess closer to days than months, but I'm not speaking for him.

oh right got email notification sorry about that,i'd be happy to spend the time, fix it? is there an way to bulk delete those associations? its fiddly trying to use the table, still trying to get my head around using freebase, havn't see an example of automatic importation of wikpedia data and making and importing lists is very slow.

i don't see the tds imported in the sandbox

I haven't found a way of bulk deleting associations myself yet. Besides the data is worth keeping, it just needs to be moved from one box to another (from the point of view of an individual politician). I'll try and do some myself.

If anyone knows a quicker way to edit a whole lot of CVTs (such as this Government Positions Held one) at once, I'd be happy to hear it.

The sandbox has refreshed since then, and I think the load of politicians from Wikipedia got abandoned.

I think the Government types are some of the most complex in Freebase; there's one help topic here, but maybe we need more?

Would individual party membership need to be date-mediated, as parties might join or leave the coalition without the coalition itself actually dissolving.

The next question is, is this just an Organization with an Organization Type of Coalition or Coalition Party, and member parties just Organization Members?

Good points. Yes, I hadn't considered joining and leaving, which certainly happens.

You're right, this could be typed as an Organization - how about "Political Coalition" for the name of the (new) Organization Type?

My other concern - which neither of these solutions addresses, unfortunately - was how to enter the winner of an election if it was a coalition (see for example the Australian election of 2004 or any recent German or Brazilian elections). I'm thinking now it should be a list of parties, but it would be convenient to be able to enter a coalition instead.

True. (And if you wanted to ask "in which elections did coalition Y form the government?", you'd need to check that all parties in Y won the election and that no other parties won that election ...)

I'm happy to stay with option 1 for now - I'll add in a list of parties for election winners, and use Organization with type "Political Coalition" for such topics.

Great, thanks.

What does that mean? Will a bot classify such topics and/or edit them? Should we edit them manually (e.g. make sure information is captured before the wrong type is removed)?

There's already a task in place to migrate the political appointment data to the Appointment schema: DA-797.

In terms of modifiying the appointment schema, I can think of two approaches: 1. Add a property to the CVT for role of the appointer 2. Add a property to Appointed Role for the role(s) which can appoint to that role. The disadvantage to two is that it might get confusing -- a lot of positions are normally elected, but can be filled by appointment if they are vacated mid-term.

And the disadvantage to approach #1 is repetition, since all the appointed will likely share the same appointer office/role/title.

Hmm, DA-797 is actually migration in the opposite direction to what I had in mind from the perspective of notable types.

So I assume the reason there are appointer/appointee/appointment types in the People domain instead of folding them into the Government domain is that there are non-government positions that are appointed that we want to capture? I'm having trouble coming up with examples outside of governments, where there are positions of power that grant authority to appointers and needs for hierarchies and clear lines of responsibility and blame that necessitate appointees. I suppose businesses can appoint leadership positions, but those are not what one typically thinks of when one thinks of appointees.

The original discussion on appointments is here: Employer?

Even if we wanted to keep governmental appointees distinct from other appointees, we'd still have to refactor to a general "Political Appointment" model, since lots of governmental positions are appointed without fitting into the Politician schema. Even omitting corporate appointments, judicial and ecclesiastical roles are often filled by appointment, so I think one general schema is much preferable to _n_ appointment schemas, every time someone finds another niche of appointments that they want to enter.

Ah, OK. I was counting judicial positions within my overreaching definition of "big government". Fair enough.

My concern lies with the weakening of the type Politician by removing properties for political appointments, thus making instances less notable as politicians through the Notable Types algorithm. So much of a politician's career is already modeled on other, more generic, types: employment history on Person, activism via Activist, organization participation via Organization Member. To say someone is an "appointer" or "appointee" would make no sense outside of Freebase.

Nevertheless, I have to admit the schema merge makes sense from a data modeling perspective; we just have to figure out a way to deal with it for Notable Types, somehow. Alas, the problem is mine, not yours.

Is there any reason not to follow the procedures and terms as defined and used by the House and the Senate?

I there any reason not to conform to the methodology of the House and the Senate?

http://clerk.house.gov/art_history/house_history/Session_Dates/sessionsAll.html

http://senate.gov/pagelayout/reference/one_item_and_teasers/Years_to_Congress.ht 

Here's the Senate's definition of session:

session - The period during which Congress assembles and carries on its regular business. Each Congress generally has two regular sessions (a first session and a second session), based on the constitutional mandate that Congress assemble at least once each year. 

It seems this schema has been improperly designed, but my knowledge of schemas is not sufficient for me to state this with any degree of certainty.

The House and Senate doc's, however, are clear, and, furthermore, are clearly different from Freebase.

What is the next step here?

Actually, in looking at this more closely, the issue is that the property "sessions" is associated with the type Governmental Body which is generic to all types of governments, not something specific to the United States. I'm not familiar enough with the world's governments to know if there's a good word which would be more appropriate than "sessions" for which is encoded here.

The next step is to get more informed opinions from someone that knows more about this than me, specifically the domain administrators http://www.freebase.com/view/government#domain-members.  They're the only ones that can change things anyway, assuming that a change is needed.

Thanks for catching the ambiguity in the terminology.  The schema is intended to capture Congresses (Knessets, Dumas, Parliaments, etc.), rather than what are typically called sessions, so name "session" was probably the wrong choice. The use seems to be that the period between elections is called by the name of the legislative body, but we can't call the type and properties "Legislative Body", since that would be too easily confused with the "Governmental Body" type.

Maybe "Instance of Governmental Body" (since executive bodies like the Swiss Federal Council also use this type), or something similar?

Please let me know what you think.

W/r/t political parties, looks like there are currently only 3:

• Unity Party
• New Democracy
• People's National Party

Then there's the Central Committee, which frankly I don't know what it is :)

I thought there were more because DNC was throwing me (thinking it was the party). That should be easy enough to clean up - luckily it is pretty bare.

Re Political parties:  I see the issue now. These are absolutely not governmental bodies (well, the Central Committee, maybe, but whatever it is, it isn't also a political party).

To the broader question, I'm going to try to sum it up, and see if you agree (or anyone else out there in discussion-land!):

The only type that allows for appointments is the Government Office Held CVT, accessible only if the appointee is a Politician.  However, lots of positions people are appointed to in government are not related to Governmental Bodies. There is the additional issue of being an employee, irrespective of how you got the job. The White House Chief of Staff is surely employed by the Office of the President of the United States, although he is appointed by the POTUS.

Here's my off-the-top-of-my-head proposal:

1. Remove the "appointed by" part of the schema from Government Office Held.

2. Create a new schema for political appointees that would record Appointer, Appointee, Date, and Position.

3. Appointees in the new schema can then additionally be typed as politicians (a la Roland Burris), ambassadors (there's a schema kicking around for that somewhere), employees, judges, etc.

This would result in a minor denormalization (all appointees would probably have their position/title and start date entered in two locations), but would save us from trying to do some insane, octopus-like model for all possible government people.

+1 for a separation of Civil servants from Politicians in the schema. 

#2. Does it have to be specific to political appointments?  It would be great to use the appointee/appointer relationship in other domains. (e.g. boards of companies etc..)

#3. skud's ambassador type is here.

Also, there is a mess between specific and general titles e.g. 'Mayor of San Francisco' and 'Mayor' which both exist in Government office and titles.  I think the title should be specific.  But I'd like to also somehow note that the title 'Mayor of SF' is a 'Mayor'.  (similar to Prince of Wales is a Prince in the Royalty & Nobilityschema)

Sprocket: I like the appointment schema; it's simple and clean.  Should appointments have a "to" value?  Or should that be left solely as part of the tenure of the appropriate cotypes (board membership, gov't position held, etc.)?  I'm asking because the act of appointing necessarily has a date of appointment, but doesn't itself have an end date.

I agree that the title should be specific. Linking the title to the role is a good idea.

@jeff.  thanks for the feedback.  I took a look at it and didn't like the overlap with the dates - I hate filling in the same info twice.  I've removed the 'to' date and also changed the 'from' date to 'declared on'.

'declared on'  will be the date the appointment was announced, even though the appointee might not take up the post until a later date.  It captures some extra data, rather than duplicating data.

Having done a bit of googling, it does seem 'appointment' in relation to a company might just be an industry term related primarily to construction.

In the construction industry 'appointment' is used when referring to the award of a contract to a firm - see this pdf for example.

The dictionary definition for appointment explicitly states person, and I'm thinking that using the schema for contracts might get messy.

I think business contracts are better in a seperate business orientated scheme, if required, so I'll change the documentation for appointment and include person as a type for appointee.

The construction "appointment" usage is also culture specific (ie Brit-speak).  In the U.S. contracts are put out to/for bid and then the contract is awarded to the bidder with the best political connections.  :-) No appointments involved.

All the U.S. instances in Iain's Google search use "tender" as a verb, not a noun (ie "tender appointment" as the opposite of "tender resignation" where the word means "offer").

I've loaded 11,623 IDs in the production database.  They can be seen in the U.S. Politician base and they can be accessed directly by ID by using URLs of the form http://uspolitician.freebase.com/view/base/uspolitician/thomas_id/K000105

I'm testing another file on the sandbox which will hopefully up the count by another 100-200, but after that the last couple hundred entries will probably require hand resolution.  The ones I've looked at so far have been things like typos in birthdates on data entry, duplicate topics, topics with no types or structured data, etc.

The notification about this reply apparently scrolled off my home page before I saw it, so I just noticed it now when I was looking at the sandbox to see if it had been refreshed yet.

I get an error when I try to look at DA-653, but since I hadn't seen any action, I created 

http://www.freebase.com/type/schema/base/uspolitician/u_s_congressperson    which has an enumeration in a namespace that I created as well as a URL template that links back to the Congressional Bioguide.

I've loaded the first 200 or so (all the As) which I had hand verified on the production database and will populate the entire list (at least the 95% or so that I can resolve) on the sandbox after it refreshes tonight so that folks can check out the rest of them.

Perhaps Metaweb can get their offshore data entry folks to clean up the remaining 5% which require hand resolution. 

This is a good point. Note that the United States topic itself, under the Country type, has a Form of government property. (Hmm… incorrectly given right now as Democracy and as Constitution.) I see you added Constitutional republic to that list.

There is an interesting dichotomy between a nation and the government thereof; it is hard to say which should really have information about the type of the government.

It applies not only to countries but to smaller jurisdictions as well… hm. Worth thinking about.

Political parties may operate across multiple jurisdictions within a country, but do not operate between countries, so I think the country is still the right level of distinction. (Movements may have parties in multiple countries, and parties may coöperate, as the Green and Communist parties, for example, but that is not the same as having one party in multiple countries.)

The cases where local affiliates of national parties have differing names, such as DFL in Minnesota or the Constitution Party in California, are the exception rather than the rule, and I am not convinced that the few cases warrant changing the model. These local groups can and should be typed as Organizations to model their affiliations and geographical scope, but the national parties should be the primary topics of interest.

Making organization an included type of political parties would solve most of these issues. State political parties would be typed as organizations that would have a suborganization link to the national party and county parties would be suborganizations of state parties. I can't think of anycases where a political party would not also be considered an organization. From a quick glance, it seems to me that all of the organization properties are applicable to political parties.

I've changed my mind on changing the country link, based on Chris' arguments, and because it would make queries on politicians much harder. If we wanted all Democratic mayors of U.S. cities, we would have to construct a two step query instead of just querying against membership in a single political party.

Close, but not quite. Freebase is internationalized: that means that every string is labeled as being in a particular language. The strings seen in the Web client view—both the names and the aliases—are explicitly labeled as English. So to say that the English name of the Russian Ministry of Defense is Министерство обороны Российской Федерации would be wrong. The Russian name of the Russian Ministry of Defense is Министерство обороны Российской Федерации, and there is a way to say that in MQL, it just can’t be accessed in the Web client right now.

Short version: for now, any words you see or enter in the Web client should be English. 

OK, I think I follow that.  Thanks for the explanation!

Unfortunately, I'm finding that not everything has an official English name, and the translation is iffy on many of those.  In those cases I am leaving it in the local language; maybe someday we'll find a good English name and we can correct.

Yeah, that's a tricky one.  Agency and ministry heads are frequently appointed positions, which could be an argument for their being typed as politicians, but pretty much everybody working under them would pretty clearly be an employee.  Maybe having the appointment information on the Government Position Held CVT is a mistake, and there should be a separate type for Appointed Official (or something).

Agency vs. Body is clearer, although maybe the documentation isn't -- a body should be reserved for legislatures, oligarchic coucils (like Switzerland's executive), and the like. Agencies are, well, agencies. (I confess that I'm not sure where departments and ministries fit in this schema, however.)

You mean you don't have a clear answer? [checks sky to see if it is falling]

In all seriousness, I forget about the appointed aspect of the Ministers/Dept heads.  It just seems strange to have a person's career not have everything in the employment history area (at least for full time positions).  Ugh, I have no idea what the right thing to do is. 

And I think Ministries are more Agencies than Bodies.   That's where I've been sticking them, so if I'm wrong, you may hear some cursing :)

As a good starting place to explore this information, you should take a look at the Filter Table View. Here's a view that shows all U.S. Presidents who served in the Army as Brigadiers General

To do more detailed searches to do things like aggregate the number of presidents who have served in each branch of the armed services you may want to look into MQL, the Metaweb Query Language.

Hope this gives you a good start.

Sorry, that link got screwed up. Hopefully this one works:

Filter Table View

Finally, this should work:

http://www.freebase.com/view/guid/9202a8c04000641f8000000008ffcc10