I'm trying do some analysis on locations where people are going during winters. The approach I'm following is get tweets from a specific city (say, New York) and with the keyword Foursquare. Then use foursquare data for that user to see his/her checkins and try to trace a pattern.
So, I'm stuck in the first phase. How do I get those tweets from ONE city and with the keyword FOURSQUARE. I'm not sure if I understood how to use streaming API correctly and the ReST API isn't working (shows NOT AUTHORISED)
Could you tell me a detailed procedure for a rookie to understand the process of doing the above mentioned process. Also, let me know if you have a better approach for analysing trends in check ins.
Thanks
You want to read these:
https://dev.twitter.com/docs/api/1/get/search
https://dev.twitter.com/docs/platform-objects/places
You can give Twitter a latitude/longitude coordinate and a radius, or you can use the "place" field as a filter. Either way, expect to fine-tune this a bit to fit your needs. You also need to take into account that a lot of people might tweet without location services enabled.
If you want to use the REST API, you need to get an API key from twitter.
Related
I'm developing an iOS application that let the user to search for a person throught the Graph API.
What I want is the SAME behavior that it's present on the Facebook website. You know when you begin to search for a person in the top text input? The first results will be mostly your friends AND some people you MAY know or people you already looked for.
The problem? Try to use the same search pattern here to search a person: Graph Api Explorer
The Graph Api returns DIFFERENT results than the search input on the Facebook website.
Does anyone knows why? Is there a way to achieve the same results?
Facebook are using many algorithms to display search result like Relevance Indicators, Complexities of User-Centric Search and The Product.
One of the algorithm to display result on their page as below.
Personal Context:
Unlike most search engines, every Facebook search involves two key elements - a query and a querier.
Just as we need to understand the query, it’s as essential to understand the person behind the query.
People are more likely to be looking for things located in their own city/country or for people who share the same college/workplace.
We consider this information and much more when ranking results. The more we know about you, the better your search results will be.
In Graph API, they are not using this algorithm.They are just displaying the queried result. Hence you can not achieve same result using graph search API.
To achieve this you can use following apporach -
Get the friend list of user using me/friends?limit=1&offset=1
Get the user list using search api
merge both the result
show result(s) to user
For more information(approach/algorithm) you can check Intro to Facebook Search
Is there a way to achieve the same results? - NO
Does anyone knows why? - NOT REALLY
(Edit: Seems in another answer, someone does actually, but it doesn't change the answer for "If you can achieve it")
But its safe to presume that Facebook does not allow all functionality through the API, why would they after all ? They need to keep the people coming to their own platform. So I can't give you a straight forward response on WHY, but IF ? Not possible, there is zero documentation about more specified search for type user. When you request user friends, you will only get the user friends who are using the same app starting v2.0
Am afraid that you will have to drop the functionality you want to achieve.
It is not just the graph search. When you refresh your TimeLine. The order of posts gets changed every time because Facebook takes a Pull on Demand approach. Which means whenever you login, the data from your friends is fetched. Which is why facebook has a limit to maximum number of friends.
Talking about the Graph search and Graph API. They are not same and the Graph Search cannot be accessed through the Graph API. So, you would have to change your approach.
To explain why the graph search gives different results on same search term. I would guess that it follows the game Pull on Demand model ( although it is not open and we cannot know for sure ). Following that model makes sense though.
Thanks
I want to get a list of users who have used a particular hashtag. Eg. #ManOfSteel.
How can I get a list of the users who are using that hashtag as well as their details (like which city they are tweeting from) using a twitter API or any other means?
Yes, you can. And it's quite simple, really.
According to the documentation:
GET search/tweets: - Returns a collection of relevant Tweets matching a specified query.
Resource URL: https://api.twitter.com/1.1/search/tweets.json
Now if you scroll down on that page, it gives an example of what a query for a tweet returns (I took a screen, sorry about the appalling arrow, it's 10am here and I haven't had my tea yet).
Great! So you know the URL and method (GET) that you need to get your data. As for searching for a hashtag specifically, the query documentation is what you're after.
This is basically as urlencoded string in the GET request like: ?q=#hashtag. Perform the search like that and you'll get back the data above. Then just loop through it, find the user object, and grab the value location key if you want the user's city.
Now, as for a library to interact with the twitter API, you haven't even stated what language you're using. From your profile, you use JAVA I'm guessing. Regardless, checkout the libraries page on the twitter dev site - there's one for JAVA that looks pretty good (and many for other languages too).
If you were using php, this post would be immensely helpful.
I have seen numerous companies doing like Twitter lotteries where users got to retweet their tweet and then one of retweeters will get the prize or whatever.
I was wondering now how do they get the list of all retweeters or pick the winner?
I checked Twitter API and found only this thing: GET statuses/retweets/:id but it returns maximum of 100 retweeters. Is that the only way?
It looks likes there's a couple services out there doing almost exactly this. A quick google pulls up http://onekontest.com/ and there's a few other Twitter contest services, but they all seem to be different levels of broken since they haven't kept up with changes to the API.
As far as the Twitter API itself is concerned, if you were expecting more than 100 responses, I think using GET statuses/mentions makes the most sense. That API call returns any mentions of a user, and you can pass the flag include_rts to include any retweets of your tweets. Then, if you wanted to list RTs of a specific tweet, you could check the in_reply_to_status_id field in the returned data to see if it matches the original tweet ID. This API call only returns the last 800 status, 200 at a time, so if you expect a bunch of data, you would need to poll the API repeatedly over time to get all the tweets. I imagine services like favstar are doing exactly this, just on a larger scale.
If you're actually looking for code to do something like this, I wrote a sinatra app called twitter-rss-digest which handles querying Twitter over time to track different sorts of queries. It's pretty rough, and doesn't quite handle this specifically, but it might point you in the right direction if you want to code something.
The Twitter API has an endpoint that returns up to 100 retweeter IDs for a given tweet.
Note that for historical reasons, that endpoint only returns up to 100 results and the documentation about the cursor is useless. Twitter refused to fix the situation.
2013 note
That said, the first thread on the Developers site that surfaced in a quick google has #episod, a Twitter employee saying:
You can't likely get to all of them. The data is spread out. Your best bet is to use the REST API to determine many of the users who performed the retweet, but the data will still be constrained.
It's easiest to track retweets as they happen rather than try to find them from the past. Use the Streaming API for that.
I like muffinista's method, but I think if you want a 100% complete list of retweets, simply enable the retweet email notifications and write a script that polls the email box for those matching the subject "retweeted one of your Tweets!" and put the data into a table. Do this right from the start.
The site https://twren.ch/ enlists all the retweeters for a given tweet (note that it only enlists retweeters who are direct followers of the source tweeter.) Nevertheless its probably the only public source available.
I am trying to get the "generic location" of a location through foursqaure.
For example - for a specific geolong, geolat provided to Foursquare API - is there an API/algorithm I can use, in order to determine if user is at home/mall/suburbs/city (generic locations)?
I could process the location type and determine this, but i was wondering if there is an easier way out there?
Since the vast majority of foursquare venues have one or more categories, you could use those to help you identify where a user is. If you have an oauth token for a user, you can access their check-in history through this endpoint: https://developer.foursquare.com/docs/users/users.html.
You'll then see on their most recent check-in a list of categories, like "home," "bar," "nightlife," "movie theatre," that you can use to get an idea of what class of place they are at.
The database actually has very few issues with misspellings, duplicates, or false positives, but you will find some issues with miscategorized homes (many college dorms, for example, classify themselves as nightclubs or strip clubs as jokes). For a majority of locations, however, the results should be useful.
Have you considered GeoNames API for searching Wikipedia by location? Details here.
Using Foursquare's API is a pretty clever idea, but based on prior experiences, you may get some false positives due to mis-spellings and/or duplicate listings.
What I'm trying to do is pull some search results, and sort them by users.
Right now I'm using
$to->OAuthRequest('http://search.twitter.com/search.json', array('q' => 'search-term'), 'GET');
to get the results, but I can't actually get information about the user like follower_count, etc. so I though I'd take each result's user_id, and run it through http://twitter.com/users/show.xml to get the rest of the information.
The problem is, if I get 50 results that way, that's 50 API calls, which is 1/3 of the hourly limit...
Is there a better way to do this?
Update:
I have requested whitelisting, and the guys at Twitter told me I should instead use the new Streaming API.
This kinda sucks since it's in Alpha testing, and the documentation is unclear, but the concept looks cool, and the limits are different than the regular API. Gonna dive into it later.
Only reliable user information you can get from twitter search results are username (from_user) and profile image url. If you want more information, you will have to call the users/show method for each user. There is no workaround or better way(that I'm aware of).
You will have to Cache user information and request whitelisting.