Discussions on Help
Start a New Discussion
-
-
Hi there,
I wonder if it is possible to include a date into the headquarter-property of the organization-type. I think it would be very helpful to describe "old" organizations like the Zionist Organisation (http://www.freebase.com/edit/topic/en/world_zionist_organization). For example, this organization changed its headquarter a few times (Vienna, Berlin, Copenhagen, London, Jerusalem)
Thanx,
f.
-
-
-
Hi there,
At the moment it is not possible to filter collections by address/country. Is this a bug?
best regards,
Frank
-
You have to use the Country property of the Postal Code type; country isn't explicitly represented in the Address type.
-
Thank you,
is there also the possibility to use an operator like "OR", e.g. Country name is "United Kingdom" or "Ireland"?
best regards,
Frank
-
There isn't, unfortunately. There's an open bug for this here: CLI-5221. Please vote for it if you want it fixed!
-
-
-
Hi there,
Last days I noticed that postal codes are not shown properly. To give two examples:
the postal code is shown as URL
- In the view: http://www.freebase.com/view/base/jewlib/views/libraries_archives_museums_and_private_judaica_holders the postal code is not shown in every case (check United Kingdom instiution, like Bodleian Library, Durham University Library etc.)
best regards,
Frank
-
The postal code value is actually stored as a property on the Postal Code type (name "postal code"), rather than the name of the topic (this is an anomalous type in Freebase in this regard). You can display this property in the table, which should help work around the lack of display-name issue.
-
Is there any reason not to populate the names of all these topics with their textual values? What was the reason for having a separate field in the first place? The ability to make it unique?
-
-
-
Hi there,
unfortunately the button next to the edit-button in topic-edit-mode does not work since two days. I already tried different browsers (IE, Firefox, Chrome) on different machines. This button is very helful in case somebody tried to import a bunch of data to a type...
Can somebody tell me what the problem is?
best regards,
Frank
-
That's a bug all right. I've entered it as CLI-9143
-
-
-
I'd like to add a list of types of Orthoses, for example "Foot Orthosis", "Ankle Foot Orthosis", "Knee Ankle Foot Orthosis", and yes even "Hip Knee Ankle Foot Orthosis" (these are usually referred to by their acronyms, AFO, KAFO, and HKAFOs). Also there are orthotics for wrist, elbow, shoulder and combinations of the above, and for back, neck, and pretty much any other body part that I missed in that list.
They also each come in different types, for example rigid, semi-rigid, soft, etc.
I have no idea how this should best be done - I'm new at this! Advice would be very welcome.
-
Good question. We might need a new property on either Medical Treatment or Medical Device to allow for subclasses/superclasses of devices. Do any of the medical admins want to weigh in?
-
Sure. Seems like a parent-child phylogeny pattern for Medical Device is unavoidable here. We will still need a property to indicate the body part a device is designed for -- perhaps on the Medical Device type itself? As for different types of orthoses, such as rigid vs soft, I'm not sure how that should be captured via a generic property. Ideas?
-
-
-
While I checked some freebase topics I noticed that not every type is shown on the frontpage of a topic. For example the Einstein topics includes a few more types: http://jewlib.www.jewlib.freebase.com/view/en/albert_einstein that it presented on the first page. I just ask because our "Originator"-type also do not appear.
cheers,
Frank
-
Hi Frank,
Client logic displays only types in the Commons on the Consumer Page of a topic (i.e. url has /view/ before topic id). All types, including user types, are displayed on the topic's Edit Page (substitute /view/ with /edit/topic/). For the Einstein page you quoted, this is:
http://jewlib.www.jewlib.freebase.com/edit/topic/en/albert_einstein
-
I think the reasoning behind this is that it's best to show only the reviewed, high quality types to new visitors.
Unfortunately that also hides good quality user types which haven't yet made it to the commons.
However, you can get the base types to be shown by appending /-/base/jewlib to the end of the url as so:
http://www.freebase.com/view/en/albert_einstein/-/base/jewlib
-
Frank, Faye is correct - but how did you get that URL by the way?
-
Masouras, good question...I just copied the address out of the browser window.
I understand, but don't you think it might be helpful if there will be a clearer sign on additional content except the "Edit and show details"-button?
f.
-
Hi there,
I wonder if there is a possibility to set a more prominent link to (non-commons)-user provided types in future. As I pointed out before, I am really not happy with the fact that user contributed data which is not contained in the common-types is hidden. In my opinion the "Edit and show details button" is irritating. Just to give one example:
Freebase-standard-view:
http://www.freebase.com/view/guid/9202a8c04000641f800000000c38d9a9
Jewlib (with /-/base/jewlib extension):
http://www.freebase.com/view/guid/9202a8c04000641f800000000c38d9a9/-/base/jewlib
cheers,
Frank
-
If this is important to you, I would open a bug report/feature request at http://bugs.freebase.com There was a bunch of negative feedback when the recent redesign was introduced, but I've never seen any indication that the designers think anything other than they got it exactly right. Having a bug report will allow you to track progress and others to vote to support your request.
Personally, I think that "show details" is much, much too well hidden. Not only is it way at the bottom of the screen so that you have to scroll down to even see it, but then it's in the 2nd half of the compound "edit and show details."
If it were me, I'd have a separate button for "Show details" (even if it went to the same page) and I'd have it at the top of the screen. This would allow me to instrument the UI to see what the users were actually selecting. If I were optimizing for the user instead of programming effort or computational efficiency, I'd only have the "Show Details" button present and/or active if there were actually additional details to be shown which were not displayed on the main topic page.
It may be that Metaweb is trying to push people to design their own applications instead of depending on the Freebase.com web client, but that, in my opinion, would be a mistake.
-
It would be great if you open a bug report/feature. Thank you!
And: Yes, I am totally with you...
f.
-
>> I wonder if there is a possibility to set a more prominent link
>> to (non-commons)-user provided types in future. As I pointed out
>> before, I am really not happy with the fact that user contributed data
>> which is not contained in the common-types is hidden.
Yes, it's a problem. We've been trying to rethink how to surface good bases but it's been a slow process as there have been a bunch of other big projects going on. It's next on the list, though.
> I've never seen any indication that the designers think anything other
> than they got it exactly right.
If that's just based on the fact that there hasn't been a design change yet, I think that's an unfair statement.
-
Jason - It's based on more than inaction. There's an extensive discussion from a couple months ago where the only people who gave any indication that they thought there was a problem were the users.
http://lists.freebase.com/pipermail/freebase-experts/2009-June/000699.html
Swapping two buttons between the top and the bottom of the page, relabelling a button with a label that makes sense to users, or even adding a new button with a separate label (but that is wired to go to the same place) are 5 minute tasks (perhaps the last is 10 minutes), so they certainly could get prioritized without having a significant impact on other tasks.
I've got a lot more to say about this, but perhaps it's more appropriate for email...
-
What about having reasonably prominent direct links to the top bases the topic is in? Would that address the concern?
-
Yes, that might be a interim solution. I personally think that it would be better to name all the types which are existent for a topic. Maybe you can distinguish between common-types and user-created-types by using different visual designs or sth like that?
Especially in the case of topics which are not described by common type it seems to me complicated that you only get to the bases which contain the types. (and the have to search again for the topic)
f.
-
@tfmorris: in that same thread, I said:
"the transition to editing from the browse version of the topic page is one of the biggest design challenges we have in terms of balancing the expectations of the wide variety of people who use Freebase. I expect us to go through many iterations of this part of the interface over the next few months."
and
"the topic page mode switch is likely to evolve a lot over the next few months, and all this feedback is incredibly valuable. thank you for it!"
I could have been more general: the topic page remains a focus of our design efforts, and no one on this design team thinks it's exactly right as it stands. we know we need to do a better job of surfacing the best data that the community has put together for each topic, and are working through several design and performance-related challenges to make this possible. -
Sorry for the delayed reply. I've been moving house and haven't had any credits in my "spare time" account.
James is totally correct. He did offer an encouraging reply to that thread. Unfortunately, it came at the tail end after I'd already tuned the thread out as not going anywhere, so it failed to impress itself on my psyche.
I'm not sure how many iterations have been done in the last two months, but I just double checked the current display rendering and I think another iteration or two could improve things more. Since the base-specific consumer page is much more nicely styled than the generic edit page, it sounds like that's an additional navigation destination that needs to be considered in the mix as well.
Frank - I think my wording may have mislead you. I was suggesting that *you* file a bug report since you are the one who is directly affected and most interested in seeing this fixed. I'm not planning on filing a bug report myself.
-
I found that thread, this thread, and others like it to be very useful inputs into the balancing act the design team is trying to perform here. look for progress specific to this thread a couple of releases from now.
-
-
-
Hi Jeff.
I'm new to FreeBase - just getting into it - researching the "Quotation" field for People - which is looks like you created. Wondering if I can ask you a question about it.My organization is about to have 1000s of people combing through news stories looking for statements that politicians have made about various topics such as envioronment, health care, budget, etc.
It occurred to me that we could contribute all of these attributions to Freebase. I'm wondering if the Quotation field is the proper place to put them. A sample entry might contain the following:
Person: Arnold Schwarzenegger
Subject: Environment
Quote: "We now know that what we've done in the past 100 years has caused such unbelievable damage to the world. We didn't know better, but now we do, and now it's not okay. There are certain things we know will happen in the next 30 to 40 years if we don't roll it back. So we have to start doing it now."
Source: http://money.cnn.com/magazines/fortune/fortune_archive/2007/04/02/8403410/index.htm
Can you point me in the right direction? Even better, to a simple API script that I could use to make these contributions?
-
Sorry I haven't been able to respond sooner. I think that Quotation is the right place for this kind of data. One possible issue is that "source" can't be a weblink -- "source" goes to a topic that represents the place the quotation appeared (in this case, probably an interview titled "The Governator's green agenda"). If you're just copying URLs, though, and not really going for the whole bibliographic nine yards, the URL could go into the "web links" property.
-
-
-
I have only a short question: I wonder if it is possible to add more than one author to a written work by using the import tool?
best regards,
Frank
-
It depends on how you're loading the data, but if you're just trying to load a bunch of book/author pairs, you can do it by duplicating the book on two lines. E.g.:
BookA / Author1
BookA / Author2
The two BookAs can be merged to the same topic, so you should end up with two authors on the same book. UNLESS you are creating a new book, in which case two book topics will be created. (Note that I haven't actually tried this, but I can't think of a reason it shouldn't work.)
There's no way to put multiple values for a non-unique property in one line, though.
-
Great! thanx
f.
-
| UNLESS you are creating a new book, in which case two book topics will be created.
@jeff Actually, you can also add new books this way and only a single book will be created. The tool matches items with the same name.
-
-
-
Only a short question: I wonder if it is possible to get statistics of bases (like these ones google analytics provide)?
cheers,
f.
-
Someone from Metaweb should provide a definitive answer, but as far as I'm aware Metaweb keeps all Freebase usage statistics to themselves. Obviously if you were going to write your own application (either hosted on Acre or running separately), you could insert whatever tracking you wanted.
-
Frank, we don't actually have per-base traffic statistics - other than the ad-hoc, top bases stats we get from google analytics. We are evaluating our next steps with Bases and whether we can expose any of that to the base owners.
I assume you want base traffic statistics rather than new topics/users - which is something you can get a good idea about from the base newsfed.
-
-
-
I'm still getting my feet under me on Freebase, but it seems to me that this page has been accidentally connected to a lot of non-Star-Wars pages (I just relinked 'Pig Bodine' to a different 'Human').
Seems like the key would be to rename this one 'Human (Star Wars)' to avoid this? any suggestions?
-
In general, we don't use Wikipedia-style parenthetical disambiguators on Freebase. The idea is that the combination of the type of the topic and the "pop-out" description which appears when linking to the topic should be enough to ensure that users link to the correct topic.
Obviously, mistakes will happen, but as they're nice and easy to fix I don't see this as serious problem.
-
Pak21: how is it easy to fix? As I understand it, I look at the list under Human (that's the Star Wars one) and I can get a list of the members of that category, but when I find a wrong entry, I have to go to *that* fictional character [click], find the Edit button [scroll, click] and scroll down to the properties, and select edit [scroll, click], delete the current Human [click, type] and re-enter Human [type, scroll, scroll, click].
or is there a faster way?
-
The current site layout with separate "browse" and "edit" pages is a new feature, and there are currently a number of bugs relating to the view switching back to "browse" mode when you'd really rather stay in "edit" mode, which makes this more painful than it would be otherwise. These will hopefully be fixed soon.
One small point: the "Add more facts" button at the top-right of the page is exactly the same as the "Edit and show details" button at the bottom-left, so that will remove a little bit of scrolling etc.
-
You know, I think these two "Human" topics should be merged. The only difference is the presence of midichlorians, and we don't make a separate species all the time for fictional characters with one thing that differs from reality. Eg, Mister Ed is a "horse" not a "horse (Mister Ed)". I can't think of any characters born with psychic powers off the top of my head, but I'm sure we'd class them as "human" not "human (with psychic powers)" or whatever, too.
Conversely, Buffy Summers is classed as species = Slayer but I think that's inappropriate. A slayer isn't a species, as it's not biological and isn't inherited. A profession, perhaps. I think that needs cleanup.
-
skud: At the moment, the dropdown list ffor the Human word puts the Star Wars version (and also the Star Trek version) much (much!) higher than the Human (fictional character) which seems to be the source of some of these misclassifications.
Merging them might make sense -- but consider that the Human (Star Trek) version actually has a lot of [fictional] content, regarding the spread of humanity and warpdrive etc, and merging that with the species of Holden Caulfield (for example) would be confusing at best. I think the Star Wars version has a similar locus for fictional content. More of a concern is that the autosuggester puts the Star Wars (and Star Trek) version very high on the suggestion lists and without an obvious clue that this is specifically for certain universes.
Could the autosuggestions be improved (ordering or extract improvements might help)?
oh, and I agree about "Slayer" not being a species but a profession.
-
@trochee: I think we should get rid of the Human (fictional character) topic by merging it into Human (real species). We don't have this split for other species, eg. Horse, so why for human?
Human (Star Trek) has a lot of fictional content *on the wikipedia page*. It mostly involves fictional events in future history, and relationships with other fictional races. Wikipedia's style is to have discursive articles that spread over multiple subject areas, or sometimes, multiple pages per subject with some pages going into more detail. Freebase's is different: we have one topic page per "thing". The "thing" here is humanity/the human race/the species homo sapiens.
If we were to model fictional events and relationships in Freebase (we don't yet), we would undoubtedly make sure that the fictional universe in which the events or relationships existed were a disambiguator, so there wouldn't be any confusion. That is, the relationship "Humans are friendly with Vulcans" would be disambiguated with "Star Trek Universe", as would "Humans are part of the United Federation of Planets".
-
I agree with most of skud's merge suggestions for humans. The Star Trek universe, for example, assumes a continuity with the real universe up to a point; the fact that there is a fictional origin for the species is beside the point.
However, in the case of the Star Wars universe, what are called humans arguably bear no kinship to Homo sapiens: it's set a long time ago, in a galaxy far away, after all. However (again), if we decide to keep the Star Wars "human" topic around, we start to run into a big problem: practically every secondary-world fantasy has a race called "man" or "human", although the majority of these worlds have no connection to the real universe. Having a separate "human" topic for every fantasy universe under the sun sounds like asking for trouble.
-
*nod* ... I personally see it is, if you had two humans, one from our world and one from the fictional univere, and a skilled biologist who knows how to figure these things out, could that biologist tell you which person was "our" human, and which was the one from the fictional universe? If there is no biological difference, then I'm gonna say they're the same species.
Another way of looking at it is, could we interbreed with the species in question and produce offspring that could also breed? All a bit theoretical I know (though I bet there are plenty of people who have considered it wrt eg. Han Solo), but not a bad mental model, especially for words like eg. Anita Blake or Babylon 5.
We should also consider these questions wrt elves, dwarves, etc. Are the dwarves in Discworld the same species as the ones in Lord of the Rings? Could Gimli and Cheery Littlebottom get it on and produce dwarf babies?
And what about vampires? Is vampire a species, a disease, a special ability, or what? The vamps in Discworld reproduce sexually, but can also sire new vamps by biting. Wikipedia says:
Vampires can create new vampires with their bite, but generally prefer not to. Given the existence of vampire couples with vampire children, they seem able to reproduce sexually. Ankh-Morpork resident Arthur Winkins became a vampire as part of an inheritance that included an old castle in Überwald and the accompanying noble title. His wife Doreen Winkins refers to herself as a vampiress, or vampire by marriage, although she is not actually undead.
I'd say that vamps in Discworld are a species, but in Buffyverse (for instance) it's more of a disease/ability thing.
-
I just realised I've argued myself into a corner. A skilled biologist could presumably detect midichlorians. Um.
Oh, look, we all know midichlorians are stupid anyway, right? *wilfully ignores them*
-
Yeah, I'm fine with having only one topic...and being willing to split it if/when someone comes up with an interesting model that requires a different set of properties for humans from their universe.
I was surprised to see you suggesting we merge into /en/human (instead of a generic fictional human topic), but I'm warming to the idea.
On the what's-really-a-species question, I think that using properties differently in different universes is appropriate. E.g. in the Madman Comics universe, there are Space Aliens who don't have an identified species...in that context, "Alien" makes sense as a species. Similarly, in that comic book context, I've listed Villian as a profession. These usages wouldn't fit for characters in a more serious universe.
http://www.freebase.com/view/base/madman/views/characters_in_the_madman_universe
-
I think the strongest contender for a distinct topic might be Human (D&D), because there's a lot of structured data that could be captured...from races they can interbreed with to gods they worship to racial bonuses. That said, no one has done it yet. If someone wants to in the future, it'll still be possible to do.
-
-
-
Hi there, I uploaded a new header image to http:/freebase.com. It worked for 2 days, but now the image shown in a very small version in the upper left corner. Can you tell me what is the problem?
cheers,
Frank
-
-
O.k. I think I understood, but how can I add a licence to the header image?
-
I've raised this on the developers mailing list, but it may be a couple of days until we get an answer due to the Independence Day celebrations in the Metaweb offices :-)
-
Until somebody with deep in the bowels edit capabilities comes out of vacation mode, you could try an experiment.
On Sandbox, could you try adding your image to the header for your base there? See if the same thumbnailing issue occurs again?
If not, you could try uploading again to your freebase base, maybe with a slight change in file name? I've looked at other bases with the same dimensions for images, and not seen the reduced to 150 in length issue there.
-
Getting a bit technical here...
Looking at the header image used for my ZX Spectrum base, I note that superalecf explicitly linked it to have a CC-by license on 1 July, which will be why that doesn't get reduced.
Playing around on sandbox, it looks like the client isn't asserting a CC-by license for newly uploaded images via the edit base page, which it possibly should be.
Frank, if you're happy to mark your image as being CC-by licensed, run this query which I think will solve your issue.
-
Sorry about all this... it was actually a mistake. This image policy change was only supposed to be for images uploaded by Metaweb staff... not all uploaded images. We'll get that fixed.
In the meantime, I do believe pak21's query will fix the problem.
-
no worries...I tried both, I run the query as suggested by pak21 and I uploaded a header image with a slightly modified filename. The query did no fix the problem.
Also, I the image was showing as a thumbnail in the sandbox. Mmmh... Just let me know when you will get that fixed.
best regards,
Frank
-
Hoo boy, complicated stuff. Back from a long weekend away and let me try and explain the background here.
Basically, there's precedent out there that says it's Fair Use to use small images (150x150) without any specific license like CC-BY, GFDL, etc., but that using higher res images may not be fair use. Therefore, we set up something in our backend so that requests for images that *don't* have a specific license will be limited to 150x150. This would apply to those that are in Freebase as "Fair Use" as well as those that were uploaded without any license/terms at all, before we started requiring it at upload time.
Most image uploads happen via the gallery thing on the topic page. But, there is one category of image that's uploaded differently: base header images. When you upload a base header image, it goes (well, went) into the system with no license attached -- none at all, just blank. That meant that all those base header images were going to be affected by the 150x150 limit, which would be a problem.
We talked about this and decided that, absent any explicit statement to the contrary, uploads on via the base admin interface are presumably under the same terms as the rest of Freebase, as described in our TOS. That is, CC-BY. So we ran a query to take all those base header images with no license terms and make them CC-BY rather than blank.
What seems to have happened here is that somehow -- nobody quite knows how -- frank's jewlib header image had picked up a "fair use" license setting somehow. This means it didn't get turned to CC-BY at the same time as all the other base header images (because the query was only looking for those with no license at all), and that's why it is appearing tiny.
The fix is to use a mqlwrite query to add a CC-BY license to that header image.
I've Cc'd alecf, who will be able to tell us what MQL is needed to do that, or, if we're lucky, he'll just be able to do it and let us know when it's done ;)
-
-
-
I got a Freebase data dump file of all of the movies info (as of March 2009) . It is basically a lot of TSV flat files. One file, called films.tsv (after extraction) is a list of over 59,000 movies, with a lot of columns for dates and director and other things that have a 1 to 1 relationship with a film, including columns for a netflix id and an imdb id. This is where I would expect to find the exact data I want. In that entire file, I did not see a single value in the "netflix id" column. The same with the imdb id column: no data for any of the 60,000 movies in there. I DID however see a blog posting (back in October?) saying that Netflix keys for over 23,000 movies were recently added to the films base. Does anyone know if it is possible to find these in one of the data dump files?
-
The TSV data dumps don't currently include any keys. If you'd like to download keys in bulk, you can use the Link Export (1.6 GB, http://download.freebase.com/datadumps/). In that file, IMDB and Netflix keys will look like this:
/guid/9202a8c04000641f8000000000009e89 /type/object/key /authority/imdb/title tt0083658
/guid/9202a8c04000641f8000000000009e89 /type/object/key /authority/netflix/movie 70053131Will that work for you? Let me know if there's anything else I can help with.
-
I think it should work, depending on the formatting within the dump files, and how long it takes me to successfully pull down a 1.6 gig file...
I'll post back when done. Thanks!
-
OK. It took about an hour to download, which was faster than I expected. It also took about an hour to unzip, which was quite a bit more.
I am afraid that I now have a 25 gig file here that I am not sure how to approach. I think i still have much reading to do on your site (or somewhere) to figure out how I can start parsing some data out of that file.
Do I need to run a graph database over here locally for this?
-
Or, let me back up. What I am really trying to get is something that would normally be returned by a SQL query that looks something like this:
SELECT imdb_id, netflix_id WHERE imdb_id IS NOT NULL and netflix_id IS NOT NULL;
Is there a simpler way to get this subset of data easily?
-
You don't need any kind of database to start using this data. If you're using a Unix-like environment, the simplest way to get at the keys you want is to directly filter it as a text-file for the key-namespaces you want using grep, like this:
grep "/authority/imdb/title" freebase-datadump-quadruples.tsv > imdb_keys.tsvIn this file, keys are represented as "id property key-namespace key-value". You could also filter by "/type/object/key" to get all the keys in Freebase.
-
That worked like a charm with the imdb data giving 19,558 records.
The netflix version returned 0 records, so I shortened the "search for" string down to "/netflix/" and got a lot of records with:
/user/hal/netflix/movie (23,929 records)
and /user/hal/netflix/role (8,397 records)
Are these the netflix records you referred to above as:
guid/9202a8c04000641f8000000000009e89 /type/object/key /authority/netflix/movie 70053131
?
-
Correlating the /user/hal/netflix/movie records with the
/authority/imdb/title records, i know have 11,189 matches -- that is 11,189 movies in the freebase dump file that have both an imdb id and a netflix id.
Does that sound about right?
-
As I write this, there are 11,182 topics on the live graph which have both IMDB and Netflix title keys, so 11,189 from the slightly out of date data in the dumps is a perfectly reasonable number.
As an aside, 19 of those topics are not currently typed as a film. The first one of those that I looked at, The Human Factor, is a book and needs to be split into two topics; similar things may well apply for the other 18. Just a small health warning :-)
-
Yep, that looks right. Sorry about the namespace confusion: /user/hal/netflix/movie is correct -- it's actually exactly the same as /authority/netflix/movie. We should probably deprecate the /user/hal name for it.
-
-
-
-
dear sir,the above named is a student-final year in nairobi school and has been sent home for fees 80000 3 weeks ago.he hails from a poor family.an organization that paid for him earlire stopped.he was the best student mariakani cc 2005 and rated among the best in your constituency.kindly assist.god bless you.diminic mwaniki-0735771981-concerned well wisher
-
-
-
This is looking really good. I see you've done quite a bit of work creating types here. Please let me know if you need a hand with anything, or have any questions/comments!
K.
-
Thank you very much for your offer. I will let you know when I am stucked with a problem.
Frank
-
Hi there,
I am just wondering if the created views will update automatically. I e.g. added one more research tool:
http://jewlib.freebase.com/view/guid/9202a8c04000641f800000000ba3da1b
Will this one appear in the following view during the next days:
http://jewlib.freebase.com/view/base/jewlib/views/online_catalogs_and_databases
or do I have to create a new view?
Thanks and best wishes
Frank
-
The answer to that question is "a bit of both". Views do update automatically, but a view you create for a base shows only topics which are marked as being part of that base. This latter process doesn't happen automatically.
You can add any topic to your base by going to the "All topics" view of the base and then using the "Add more" button, or by manually adding the "/base/jewlib/topic" type to the topic itself. I've done this for the new topic you mentioned, so it should now appear. (Note that if you add a topic to a view by using the "Add more" button from that view, it will automatically satisfy the criteria to appear in that view).
Hope that makes sense - the whole "base topic" thing is confusing, and I believe is planned to be replaced "soon".
-
Thank you for your answer,
I understand. The problem is, if I add a few topics to a view,e.g. to
http://jewlib.freebase.com/view/base/jewlib/views/jewish_studies_research_collections_topics
the e.g. rel. research tools do not appear in:
http://jewlib.freebase.com/view/base/jewlib/views/online_catalogs_and_databases
It is really to hard to add each topic manually to the different views...
If I use the Add more button at the "All topics"-view I am afraid of losing the overview...
-
I think I understand your issue, but I'll run through it just to check we're talking about the same thing :-)
- You create a new research collection
- You edit the "metadata available through" property of the new topic, creating a new online research tool.
- The new online research tool doesn't appear in the second view.
If this is the issue, I'm glad to say it's easy to fix! Go to the schema for online research tool and add "/base/jewlib/topic" to the set of included types at the bottom of the page: this will mean that anything which is (in future) typed as an online research tool will automatically also acquire the magic "base topic" which makes it appear in your views. [ You may want to consider doing this for your other types as well ]
If your issue is something else, let me know and I'll see if I can help with that.
-
Thank you very much!
-
O.k. only one question is left: If somebody uses the "Research Collection"-type to create new topics, do these topics also appear in my views? This might be confusing, because not all research collections are relevant for Jewish Studies (as you could imagine).
best regards,
f.
-
Hi there, I have some problems in understanding what is going on. I will try to summarize.
I created a few types during the last weeks. Among them is the "Jewish Studies Research Collection Type":
http://jewlib.freebase.com/type/schema/base/jewlib/research_collection
The auto-fill function was fine up to now. For example when I added research collections to a specific Judaica Owner - see e.g.
http://jewlib.freebase.com/view/en/national_library_of_poland
the Owner also appear automaticly in the "research collection held by"-field
But after I made some changes to different properties of the type, it does not work anymore and/or the tie between the types Jewish Studies Collection and Judaica Owner is not shown on "Judaica Owner"-topic views? (There are some topics which keep the field filled, like:
I have no idea, what to do. (Maybe it is no real problem, because the return link still works)
best regards,
Frank
-
Weird. At a quick glance, I can't seem to locate the problem. Can you link to a pair of topics where the link is broken?
-
Oops. Typelibrarian is me; I forgot I was logged into that account.
-
The Link is not really broken. Only the e.g. owner of the collection which has to appear in the field "collection held by" (I think) is not shown, like in (Just click the edit button under type Judaica Owner/Research collections:
http://jewlib.freebase.com/view/en/yivo
http://jewlib.freebase.com/view/guid/9202a8c04000641f800000000ba4662b
http://jewlib.freebase.com/view/guid/9202a8c04000641f800000000ba4664f
Non Judaica Owner (e.g. Originator) topics show the Judaica Owners in the field "collection held by, like:
http://jewlib.freebase.com/view/en/alter_kacyzne
or
http://jewlib.freebase.com/view/en/general_jewish_labor_union
The only problem here is that the headline of each field is not shown.
best regards,
Frank
p.s. At the first moment I was suprised about your human-like-answering bots ;)
-
The second issue, with the header not showing, looks like a bug. I'll file a ticket for that.
The first issue is not really an issue: you're seeing the "collection held by" property in the edit window because that property is non-unique. Nothing is displayed in the table view because there are no instances with multiple values, and the value that is present is the current topic. If you have a collection that is owned by more than one owner, the other owner(s) will appear in the table.
-
-
-
-
I was trying to create a type schema with an "organizer" property. I want to specify that the "organizer" may be of existing type "person" or "organization". I could not figure out how I can do that. Any idea?
-
You can’t, for a couple of reasons. First, if you add a previously-unknown organizer, what type should Freebase make it? Second, if you want to make a reverse property for organizer (organizations organized, perhaps), what type should it be on?
The thing to do in this case is to make a new type, probably called Organizer. That way the Person or Organization that did the organizing can have a reverse property easily.
-
I see.
It is a common case to have union of types as the range for a property. I think that creating new types in this way may result in many types.Anyway, thank you very much.
-
teasdf
-
-
-
I really think that the people at http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.24.y.git;a=blob_plain;f=CREDITS;hb=HEAD should be added to the Software > Developer list; but this would need more than 400 new topics, with only Name, Email, Web address, PGP key id and finerprint, contribution to the Linux kernel, and snail mail address. Any help on this? Here are some UNIX scripts that act on the file above to get diferent information:
head credits #This is informational stuff, like the fields.
cat credits | awk '/^N:/ {print $2, $3, $4, $5, $6, $7}' # Print all names in alphabetical order. Replace 'N' by 'E', 'W', 'P', 'D', or 'S' for the respective field (see `head credits`).
cat credits | awk '/^N:/ {print $2, $3, $4, $5, $6, $7}' | awk '{c += 1} END {print c}' # The number of people in the file.
-
Hi andresj,
I find the easiest way to get started with this type of task is to create a saved view to define the information that you want to enter. In this case it would be Software Developers who have contributed to the Linux Kernel. If you haven't already, you should check out the tutorial on saved views to see how I made this view.
Once you have a saved view, you can click on the "Add More" button which will pop up a box with a link at the bottom that says "Go to the import tool". Click on that link and you will be taken to a tool that lets you copy and paste a list of topics (the names of the contributors) to be addded to Freebase as part of that view. I find its best to try adding them in batches of 100 to make sure they don't overload the system.
The list import tool will try to help you find any existing topics that match the topics that you're adding so that Freebase doesn't get cluttered with duplicate topics. Then when you're finished adding the new topics, they will all be automatically typed as Software Developers and they will all be added as contributors to the Linux Kernel.
The other data will have to be entered by hand or with a custom script using the Freebase API. Freebase doesn't currently store email or snail mail addresses for people so the only other data that you could add from that file would be websites and places lived (the country from the snail mail address).
Give that a try and let me know how it goes.
-
-
-
The concept of a Student Organization is not unique to Carnegie Mellon University, just as a student publication is not unique to any particular school.
How do we generalize this topic so it can be used by other schools?
-
The semantics of this type has merit, but most of these properties are covered by either the organization type or the topic type. I'll create a task to model a student organization type that includes the organization type and has a property to link an educational institution and a property to link an adviser. What other properties would be useful on this type?
-
DA-544 has been filed to track this issue. If you have suggestions for other properties, please add them as comments to the task.
-
-
-
Hi,
This query brings back expected results in MQL editor [sandbox btw]:
[
{
"/music/composition/composer" : [
{
"id" : "/en/johann_sebastian_bach"
}
],
"id" : null,
"name" : null,
"name~=" : "parti*",
"sort" : "name",
"type" : "/music/composition"
}
]same query in app editor brings back two records and quits.
Thoughts?
-
PART 2 ---
OK, the results diff cuz MQL editor was pointing to sandbox
How can acre queries point to sandbox (or can they)?
-
you can use the sandbox appeditor http://sandbox.freebase.com/tools/appeditor
-
-
-
Example with this query
http://www.freebase.com/view/en/arnold_schwarzenegger
How I can gain the description of Arnold Schwarzenegger by a MQL query:
"Arnold Schwarzenegger is an Austrian-born bodybuilder (Mr. Olympia), actor (The Terminator), and politician (Governor of California)"
-
Hi there,
you might want to check out this discussion on the developer's email discussion list:
http://lists.freebase.com/pipermail/developers/2008-June/001753.html
The basic answer is: you can't because the description is stored in a file system rather than in the graph database, which is what MQL queries. If you want to get the description you have to do it by means of the API trans service. Here are the details for that:
http://www.freebase.com/view/guid/9202a8c04000641f8000000006cf5605
You'll need the GUID of the content that you want to retrieve.
-
Thanks philg, but I think everyone likes retrieving description that is included in MQL query rather than using trans service or re-quering wikipedia.
It's my opinion.
-
Opinion noted. Meanwhile, you still need to use the trans service.
-
I agreed. Everyone wants the description. I get that it is painful to put that in the tree but geez that would have been a lot more convenient...
Don't take it wrong, but learning first how to use MQL ( ok, I get it, you don't like sql... ) and then learning that it is not what you have to use...
I bet you have that sort of topic every two weeks !!
-
Another feedback on the documentation :
Use a FAQ and a lot more examples of queries.
Examples are fast to understand. A lot more faster than everything else.
Just a page with 30 typical queries with description, image and stuff like that. That would be cool.
-
There are some MQL query examples at the bottom of the queryeditor.
-
Useful, but hard to find !!!
I had no way to find these without knowing it.
It should at least be put on top
-
-
-
I'd started using freebase today and chose to make a list of the numbers of the electors each USState send to the college. But how do I model this?
My first approach would be to specify the number as a new attribute to the Governmental Organization that the USStates are. But how?
Is this incorrect? Is there a better approach?
-
Welcome to Freebase!
Usually the best way to approach this is to create a very specific type in one your own personal domains and try adding it to some topics. For example, I've taken a crack at it here in my Politics domain.
Once, you have some data modeled using your personal type then you can contact the owners of one of the data commons types like Governmental Body and discuss whether your type can be merged into one of the more general types. For something a specific as electoral college votes it might be better off as a US-specific type.
-
-
-
Howdy Guys,
Haven't been a member long, but I was planning on doing a main domain (not a personal one, but under Special Interests). It was like automotive, but for Cycles (bikes, trikes, tandems - etc).
There are a few personal domains around that are similar. Cyclocross (Mikeshwe) and Bicycles (Collord) - I guess this would be a best of both worlds, like the automotive one.
Am I just being silly? Should I start a personal domain first? I guess this is just to find out what I can actual do, thought it would be better to ask.
With Thanks
james
-
Hi James,
You're correct - the best way to start is to model Cycling in a personal domain and post a request for review, or for harder data-modeling questions, you can join the data-modeling list (here).
My question back to you is, can you generically represent most or all bikes in a single cycling domain? It seems that mountain bikes have different specs that people are interested in (hard-tail, full suspension, type of suspension) vs road bikes (possibly materials and weight). You might want to think about domains centered around a specific biking style.
-
Shoot, here's the link to the mailing lists.
-
-
-
Is there a protocol for split-awardees for Nobel prizes? Right now, the Award Winner field on a page for winner A seems to be filled in with the _other_ award winner name B.
I would suggest the two names listed, for example
Name A (1/2)
Name B (1/2)
-
This is a general case for non-unique properties on compound-value types. The person whose topic you're on is assumed to have won (else the prize wouldn't show up), so any additional names reflect people they're sharing the award with. If you enter the award from the person's topic, you can see that that person's name is automatically filled in, and that you can add additional names below it. But I agree that it can be confusing sometimes. If you view the award from the award page, you can see both listed together -- see the 1902 and 1903 Nobellists for Physics, for example.
-
-
-
Please don't merge, there is a big confusion here: Ballet (as a dance form) was marked as a "musical genre" by somebody else at the time I flagged this for merging. These are 2 separate things, Ballet (as dance form) and Ballet (as a musical form) and THIS is a musical genre, not the dance form! I tried to separate the two but I don't know if I succeded. In any case, please don't merge. Bryan, if you see this, be so kind and remove the merge flag.Thank you
-
-
-
Good morning, Bryan
there is something very wrong on my L. Ferrero page at Composer-Works composed-Storie di neve. The computer would not let me save/store new photographs and the information I wrote at "Add description." Could you, please, have a look? Thank you, and let us hope we can save Munich from deletion!
-
Please post about problems on the actual affected topics, not randomly selected users. It is much easier to find the problem that way.
The problem with “Storie di neve” was that somehow it was a Composition but not a Topic. The article and pictures are properties of Topic, so they weren’t showing up.
How did you create “Storie di neve”? There may be a bug in our system; if you can tell us how you created it, we can correct the problem.
-
Cris, I am almost sure it is a bug.
I created it normally, as a work in "Works composed." I put a YouTube video, filled up the spaces, uploaded the picture, added a short description and then I saw that nothing was being stored except the video. I had this problem before when I wanted to correct the description on the main page. The computer is not storing what I do...
-
Cris, now I got back the pictures but the short description is lost. Never mind, I will write it again and see what happens. Thank you very much indeed
-
Storie di neve is working perfectly!
I have a similar problem at Opera composer - Opera composed, please...
Computer is not storing new information
-
Chris -- the type /music/composition does not have /common/topic as an included type. I have no idea how this happened, but I have seen one or two other types where topic was missing.
-
This was by design in the old age of primitive conservation. I have made /common/topic an included type of /music/composition, and am running a script to garden existing instances.
-
Thank you all.
I would like to add that I never write randomly. Bryan helped me the other day with an enormous problem I had in a matter of seconds!! I thought he was a bug expert and that's why I wrote to him directly.
Bryan, let us hope we save Munich and PARIS as well...
-
Chris already diagnosed and remedied the problem before I even had a chance to read this post - he's that good. I am checking on Munich and Paris - they have been flagged, but I don't see them in topics for voting. I'll keep you posted on their status.
-
Thank you. It is incredible that someone would want to delete them!
I am jolly glad that they are not on the delete list.
-
-
-
I've been slowly adding the historical results of baseball's world series, and in thinking about the 1994 world series, I hit upon a bit of a hurdle: this world series didn't occur.
As a piece of information, this fact seems relevant to me. However, in terms of integrating the "non-event" into the site, it really doesn't fit at all with the structure of all the other world series results, namely, "winner, loser, score", etc.
I guess that my question, then, is how is one to deal with non-events that are nevertheless relevant precisely because of their exceptionalism? The expectation of a world series in 1994, given the pattern set both before and after the cancelled season, separates this instance from all others in its class. However, since the actual event never occurred, I’m not sure if representing it in the current schema is either (i) possible or (ii) desirable.
There are other examples of this kind of expected-but-never-occurred event, notably in sports, and I was wondering if there are any ideas/suggestions on how to represent this type of information.
Thanks-
That's a good point. Just because an event doesn't appear in Freebase doesn't mean you can assume that it didn't occur.
I would model this by co-typing the 1994 World Series topic with a Canceled Event type and then maybe add a property to say why it was canceled.
-
Ditto what Shawn said; a new type seems the best way to handle this. I would also strongly recommend bringing up this discussion on the data-modeling mailing list.
-
cool, thanks for the feedback. I will definitely bring this up on the list.
-
-
-
Considering there's no data in it, I think that this was accidentally typed /people/person. It should probably be removed.
-
I see skud already removed it. Please keep in mind that Freebase is a community-edited Web site; that means that the community—meaning you—can and should edit it when you find problems. Any logged-in user can remove a type from a topic by using the little drop-down arrow under the type’s name.
-
I had tried it, but for some reason the change didn't go through. Thanks, though, both of you. ^_^
-
-
-
The best place for watching the total Solar Eclipse in China is Xinjiang Uygur Autonomous Region . The unique astronomical phenomenon will occur on the 1st August 2008 in China. There will not be a longer total solar eclipse until 2009 in China, so don't miss the opportunity to watch the total solar eclipse in China.On August 1, 2008, a total solar eclipse will occur as the new moon moves directly between the sun and the earth. The moon's umbral shadow will fall on China. For planning the total solar eclipse trip including transportation,hotel booking, car rental ect. just email us and we will help you with your xinjiang tour for viewing the total solar eclipse in China.
Our total solar eclipse tour begins and ends in Beijing, China . It has been reported to us that 1.5 million domestic Chinese travellers and 750,000 foreigns are expected during the Olympic period. Hence we have put our most popular optional tour ahead of the eclipse package. However, the basic and post tour are nevertheless impacted by the peak Olympic pricing period.
July 28 (Monday): ARRIVAL IN BEIJING
Arrive in Beijing on your own. After completion of customs formalities, you will be greeted by our professional tour guide holding a sign reading "total solar eclipse". You will then be transferred to your accommodations provided at the Dongjiaominxiang Hotel.
Dongjiaominxiang Hotel Beijing is a four-star hotel located in the center of the Beijing City. It is in the former embassy area. You may walk to Tiananmen Square and the Qianmen Commerical area in ten minutes. Wangfujing Commerical Street is also ten minutes walking distance. The Dongjiaominxiang Hotel in Beijing has 218 guestrooms and all rooms are equipped with air conditioning, satellite TV, cable TV, IDD, DDD and internet access. Overnight Beijing.
Of all the tourist sites of potential great interest to eclipse observers might be the Gate of
Dispelling Clouds. It is located a the Summer Palace (part of the post eclipse tour). However, if you are superstitious, it might not be a bad idea to make a small pilgrimage there before Eclipse Day. Twelve Taihu rocks depicting the twelve Chinese zodiac animals and a pair of bronze lions are in front of the gate. The lions are auspicious animals and in Chinese culture are believed to ward off evil spirits. The twelve rocks are shaped like the rat, ox, tiger, rabbit, dragon, snake, horse, sheep, monkey, rooster, dog and the pig. July 29 (Tuesday): BEIJING
Beginning today, breakfast, lunch and dinner will be included daily. After breakfast, there will be a welcome briefing in the hotel.This morning you will be among the privileged few to enjoy a behind-the-scenes view of Hutong, the old city of Beijing. Setting out with your guide on traditional three-wheel cycles, you will visit the 15th century Drum Tower for an overview of the intricate labyrinth comprising the district. Then, walk along its ancient alleyways and visit both typical quadrangle's where ordinary citizens live and gracious old estates once owned by the capital's aristocracy. You will enjoy lunch with a local family and an afternoon of shopping at the Silk Alley and Rainbow Market.
Overnight at the Dongjiaominxiang Hotel, Beijing (B,L,D).
July 30 (Wednesday): DUN HUANG
Transfer to the Beijing airport for your flight to Dunhuang (Gansu Province) in western China. Upon arrival you will be met and taken on a tour of the Mogao Grottoes and the Museum. The Grottoes are one of the most important ancient sites of Buddhist culture, preserving nearly a thousand years of cave-temple architecture, clay sculpture, mural paintings, and manuscripts dating from the 5th to 14th centuries. They are carved into desert cliffs overlooking a river valley. There are a series of 500 caves, 2400 painted sculptures and 45000 square meters of frescoes. Following your tour you will be transferred to your hotel (either Don Huang Guest House or Dun Huang International Hotel depending on which group you are in). Lunch and dinner at local restaurants. Overnight Dunhuang (B,L,D).
July 31 (Thursday): DUNHUANG TO HAMI
After breakfast, travel by motor coach from Dunhuang to Hami (Xinjiang Province) approximately 420km (250 miles) which is about a 5-hour drive. The town of Hami is a small oasis at the remotest end of China. Known in the past as Kumul it was once an important caravan stop on the northern route of the Silk Road. Accommodations provided at the Hami Hotel. Dinner and overnight in Hami (B, L, D).
August 1 (Friday): ECLIPSE DAY!
Your choice of observation of the eclipse from one of three areas. For those who want to observe totality only, there will be two centerline sites to be announced later. At both locations from 1m50 to 2m00s may be visible. The sun will be situated about 20 degrees above the western horizon for this rare alignment of the moon and sun.
The third area is very near our base site of Hami and is defined by the southern limit of the eclipse. About 30 seconds of totality is visible at the Hami city site, but Baily's Beads should be very prolonged and the purpose there will be to videorecord the Beads as part of an ongoing study of changes in the solar radius! Overnight in Hami. (B, L, D)
August 2 (Saturday): HAMI TO DUNHUANG
After breakfast depart by motorcoach to Dunhuang. Check into your hotel (Grand Sun Hotel or Dun Huang International Hotel). Afternoon tour of Mingsha (Singing Sand Dunes) Mountain which is celebrated for the sounds of the moving sand. Named Sha Jiao Moutain in the early East Han Dynasty, it is 800 square km in size with ridges and mpeaks known to be as sharp as a knife. Stepping on the sand may change its shape temporarily but the sand always returns to its original shape the next day. Walking on the sand, one can also hear a sound that could be as light as a bamboo instrument or as heavy as thunderous drums. Also visit the Crescent Spring (Yueyaquan) that is so named because of its large crescent shape. The spring has a depth of 3 meters with sweet, crystal clear water. The sand never gets into the spring and the water is never muddy or dried up. There are three treasures to be seen here: the iron back fish, the seven star grass and the five color sand. Lunch and dinner at a local restaurant. Overnight Dunhuang (B,L,D).
August 3 (Sunday):
After breakfast depart for the Dunhuang airport for your flight to Beijing. End of land tour (B).
TOUR PRICE INCLUDES:
- accommodations in double/twin bedded rooms in 3 - 4 star hotels or similar category
- bus transfers to/from solar eclipse site
- solar eclipse briefing
- round trip transfers from Beijing airport to hotel in Beijing provided you arrive and depart with the tour group.
- round trip flights between Beijing and Dunhuang, China
- all road and other transport as indicated in the itinerary
- airconditioned bus transport
- entrance fees
- sightseeing per the itinerary
- baggage handling at all hotels
- meals as shown where B=breakfast, L=lunch, D=dinner
- beverages with lunches and dinners
- local English speaking guides
- bottled water on the buses
- Hotel V.A.T (Value Added Tax), city tax, and service charges
TOUR PRICE DOES NOT INCLUDE:
- International airfare
- Any additional charge(s) and/or special documentations to be filled for importing/exporting telescopes (if even required)
- cost of visa to China
- cost of required vaccinations
- travel insurance (highly recommended)
- roundtrip Beijing airport to your hotel transfers IF YOU DO NOT ARRIVE/DEPART WITH THE GROUP
- beverages
- departure tax from China at end of tour
- optional excursions
- tips and gratuities
- personal expenses and anything else not included in the 'tour includes' features list
-
-
