Find a place description in Freebase using latitude and longitude? - latitude-longitude

I'm receiving lat and long data from Google Maps and I want to send that along to Freebase to get a description of the relevant place.
There's a geocode schema (http://www.freebase.com/location/geocode?schema=) but it seems like you have to to enter them separately?
I'd like to, let's say using the lat / long of Berlin, get the following description from Freebase: http://www.freebase.com/m/0156q
Would love any help, thank you.

If you convert your point to a small bounding box, you could get the Freebase data for all topics included in the bounding box like this:
[{
"id": null,
"name": null,
"/location/location/geolocation": [{
"latitude>=": 52.5,
"latitude<=": 52.6,
"latitude": null,
"longitude>=": 13.38,
"longitude<=": 13.39,
"longitude": null
}]
}]
You can make your bounding box arbitrarily small, but remember that for something like a city, the point probably represents something like the centroid while the actual bounds will be much larger. If you expand to the city limits, your search will, of course, return all the churches, stadiums, pubs, etc which are located within the city. You can add additional filters to only return certain types of entities if that helps for your purposes.

Related

REST API and GET with structured data

I'm designing a REST API where given an address, I will return the timezone corresponding to the address. I started along this path,
GET /api/TimeZone?address=
Now this address is a free form address field that I need to parse on the server.
To avoid errors, I'd really like to have the user send in a record like {"city": "", "state": "", "country": "" }
But I can only do this using a POST or a PUT endpoint => which semantically implies that the data is changing on the server but its actually NOT.
What would be a good way to address this?
You can have multiple fields in your query:
TimeZone?city=London&country=uk
Or if you could use hierarchical URLs such as:
TimeZone/UK
TimeZone/UK/London
I would expect the first of these to supply a bunch of cities, such as those found in most clocks, for UK, London would be the only suggestion as the whole of the UK is "London Time", but for America, there would be many suggestions.

Find relation between two entities in Freebase

I am new in Freebase and I have a simple question . I would like to use Freebase KB to find relation between two entities. For example if I have name entities "Washington" and "United States" , I would like to send a query to Freebase and get :
Location/Location/Capital or Null in the case of No relation.
Thank you very much.
If you only want to go one ply out (ie nearest neighbors), this is pretty simple to do using the reflection API if you're using the online version of Freebase. If you're using the bulk downloads, you'll need to work with whatever query engine you're using (probably SPARQL unless you converted the RDF to something else).
If you want to find the shortest path(es) regardless of who far apart they are, it becomes a graph search algorithm.
EDIT: If you only want to find capitols, you can fill in your IDs in this query:
[{
"type": "/location/administrative_division_capital_relationship",
"capital": [{
"id": null
}],
"administrative_division": [{
"id": null
}],
"limit": 1
}]
Note that for Washington, D.C., this will return null because the data isn't in Freebase.
If you need to handle arbitrary properties, you'll need to use reflection. See https://developers.google.com/freebase/mql/ch03#reflection

How do I return filtering meta data in a REST API search query

I'm currently designing and implementing a RESTful API in PHP.
The API allows users to search for hotels.
A simplified example of the search request is:
GET hotels/searchresults?location=<location> #collection of hotels within location
The response also contains some meta information about the returned collection.
The basic structure of the response is:
“meta": {
“totalNrOfHotels": 100,
"totalNrAvailable": 80
},
“hotels": [
{
“id": 123,
“name": "Hotel A"
},
{
“id": 135,
“name": "Hotel B"
},
...
]
This resource also supports pagination:
GET hotels/searchresults?location=<location>&offset=0&limit=20
Now, there are a few filters that can be applied to the search results, e.g. stars, rating score.
For example, if I want just 2 star hotels, I can query:
GET hotels/searchresults?location=<location>&offset=0&limit=20&stars=2
Now, in the user interface for filtering, it is common to display the number of options available per filter setting:
In my opinion, these numbers can be seen as meta data about the search query. So, we could add an extra field to the meta in the response:
“meta": {
“totalNrOfHotels": 100,
"totalNrAvailable": 80
“filterNrs": {
"stars”: {
“1": 1,
“2”: 9,
“3”: 39,
“4”: 12,
“5”: 11,
“none”: 9
}
}
},
“hotels": [
{“id": 123,
“name": "Hotel A"
},
{“id": 135,
“name": "Hotel B"
},
...
]
So, I have two questions:
Should this “filterNrs” property sit in the meta section, as proposed above? To me, it doesn’t make sense to be a separate resource/request
How can we deal with the fact that this can slow down the query? I’d prefer to make the “filterNrs” field optional. We are thinking of using a “metaFields" parameter to allow the user to specify which fields in the meta she would like to recieve. We already support this for the hotels returned, with a “fields” parameter. (Similar to: https://developers.google.com/youtube/2.0/developers_guide_protocol_partial). Alternatively, we put this field filterNrs (or the full meta info) in a separate resource, something like hotels/searchresults/meta. From a developers perspective would you prefer to have this split into multiple resources or have a single resource with the option to show full or partial meta information?
Does the number rated per star count varies? For example, do I get different "filterNrs" for the queries below?
GET hotels/searchresults?location=1
GET hotels/searchresults?location=2
I would expect such filters to be contextual, so different locations would return different numbers per star count, which indicates this is some form of contextual information related to the query.
Otherwise if the results are global this indicates it's a separate resource. If it's a separate resource scenario, you can use links to access the numbers and other details about it:
“meta": {
“totalNrOfHotels": 100,
"totalNrAvailable": 80
“filterNrs": {
"stars”: {
"options" : ["1", "2", "3", "4", "5", "none"],
"details" : "http://example.com/stars"
}
}
},

LinkedIn API people company

I would like to show the people numbers of a company. For that I display a result of a search.
I got a numResult but this number is not the real number of the company
IN.API.PeopleSearch()
.fields("id", "firstName", "lastName", "headline", "pictureUrl","location","public-profile-url")
.params({
"company-name": "Google",
"current-company": true,
"count": 8,
})
.result(function(result, metadata) {
setSearchResults(result, metadata);
});
}
Then I have result.numResults who is not the good number of the company. Someone has an idea?
I guess this is the number of the company about the user when he is connected.
You can retrieve specifics about companies, including the company size, from the Company API. The specific field that contains the size range for the company is employee-count-range.
In terms of accessing the data via the JavaScript API, consider using the IN.API.Raw() method.

How to get ids from Freebase given part of a name?

When I go to Freebase.com and search for bush (for example), I get suggestions like George Bush, Kate Bush, George H.W. Bush etc.
How can I get that list as IDs from the query api?
I am trying to get the a list, like this, of ids for a particular name:
[{
"id": null,
"name": "bush",
}]
You can use the search API
https://www.googleapis.com/freebase/v1/search?query=bush
or if you want to use MQL, use can use the contains operator (~=)
https://www.googleapis.com/freebase/v1/mqlread?query=[{%22name~=%22:%20%22bush%22,%20%22id%22:null}]
The contains operator does whole word matching by default. If you want partial word matches, you can add additional wild cards, but you risk running into timeouts, particularly for leading wildcards.
https://www.googleapis.com/freebase/v1/mqlread?query=[{%22name~=%22:%20%22*bush*%22,%20%22id%22:null}]