facebook extractor by date..??

[cingcpirit]

How to get topic from facebook extractor sorting by date..??
please help me..

[akivela]

Hello Cingcpirit

I didn't quite understand your question. Could you please describe the question/problem a bit more. I figured out you are using Wandora's Facebook extractor and want to sort something by date, right.

Kind Regards,
Aki / Wandora Team

[cingcpirit]

yess....i'm using facebook extractor...and i want to get data from page (like coca cola), but i want to get data in that page within a certain time, example : i want to get photo from "2 may 2012"

i'm try to using this graph :
1) https://graph.facebook.com/Koran.Fesbuk ... 2000:00:00
and
2) https://graph.facebook.com/Koran.Fesbuk ... =yesterday

but when i using facebook extractor i'm still get the photo from the past of that date....could you help me..??
i'm confuse...it was wrong from wandora or facebook...because when i type graph number 2 in browser, i got data from that date....

[mataharisenja]

Hi there..

I got the problem too.. I want to get data from facebook page by date,,
I know in twitter extractor there's a field which state that we can get data by date,
does facebook generator allows too?

[akivela]

Hello Cingcpirit and Mataharisenja

Yes, Wandora's Facebook extractor indeed lacks support for the until and since fields documented at [1]. We have to develop Wandora a bit further to allow date filtering in searches (specifically in the Search tab of Facebook graph extractor).

If you want to use until and since fields in Facebook extractions before the official fix-release, you can use the URLs tab in Wandora's Facebook extractor. Enter a complete Facebook graph URL such as https://graph.facebook.com/Koran.Fesbuk ... =yesterday (Cingcpirit example 2) into the text area. And start extracting. The extractor takes your URL as such and tries to fetch and parse the resulting Facebook JSON.

Also, if I understood the Facebook graph API documentation right, the created_time field can not be added as a parameter into the Facebook graph URLs.

Kind Regards,
Aki / Wandora Team

[1] Chapter Paging at http://developers.facebook.com/docs/reference/api/

[akivela]

I just tried to perform a Facebook graph extraction with your example URL (1) and it looks like the result set contains too much information i.e. the extractor doesn't respect given until and since arguments. I have to investigate application's behavior a little more detailed.

Kind Regards,
Aki / Wandora Team

[cingcpirit]

thanks akivela....

allright, i hope i can get data ( comment,like,photo) from page (Koran.Fesbuk) by using date...like Twitter extractor...
example : i want to get data from 2012-04-30 until 2012-05-01..using facebook extractor ..can you give me some explanation and solve my problem...?

i see twitter extractor can do that,,,like this picture :




And last Question...why when i get data using facebook extractor, it's taking to much time...i try to extract from page (Koran.fesbuk), i got 7 hours, but it didn't complete...i stop the process...is there another way to save the time and so the extractor can complete the process more quick...?? why twitter extractor do that very quickly than facebook extractor....?

Please help Me...

[akivela]

Hello

I have today (2012-05-07) updated Wandora application. Updated version fixes some issues in Wandora's Facebook extractor. To get the updated application version, just download it in our download page at http://www.wandora.org/wandora/wiki/ind ... e=Download .

Old Wandora version used to follow paging URLs of the Facebook graph feed. Paging URLs simply express how to get the next (and previous) set of results. Wandora fetched and processed these additional pages too. This page-following-feature also caused Wandora to ignore the since and until parameters and made the Facebook graph extractions very slow if the feed had many pages. Updated Wandora doesn't follow previous and next links in Facebook graph feeds.

If you want to get data from page Koran.Fesbuk using date, I suggest you to use the URLs tab in Wandora's Facebook extractor. The URL address

Code: Select all

https://graph.facebook.com/Koran.Fesbuk/albums?since=2012-04-30&until=2012-05-01&limit=100

should work. Notice, default limit is 25. If you expect and wish to receive more results, add explicit parameter in the URL. Also notice, Facebook probably has a maximum limit, you can't exceed.

Should you have any other problems with the Wandora, do drop a line.

Kind Regards,
Aki / Wandora Team

[cingcpirit]

Thank U so Much....
wandora works 100% with parameters "since" and "until"

i have try your link..but, i have new problem...when i try to get albums , why i just got 2 comment...? i want to get all of comment from each photo of album...
i have try to modified that graph(link) like change "albums" to "feed", "posts", "photos" and added limit=1000, but had no effect...
all i want is get all of photo , comment, like, from that page...

and can you give more explanation about depth...? how it works?

thanks

[akivela]

Excellent, you got the extractor working with since and until parameters.

About missing comments. You could try increase the depth value from 1 to 3. Depth value defines how far the Facebook graph extractor proceeds extraction in the graph. Value 1 means that Wandora will extract only given graph node. Graph node is specified with the Facebook identifier or Facebook graph URL. Depth value 2 means that Wandora will extract given graph node AND directly linked nodes. Depth value 3 means Wandora will extract given graph node AND directly linked nodes AND their linked nodes. In this specific case the first node is a Facebook photo album, second node is a photo in an album, and third one is a comment of the photo. Thus, I think you should use depth 3 to get also all comments of the photos if you use the URL I suggested in my earlier post. Notice my warning, I assume the depth 3 gives you also nodes that you are not interested in and nodes you have no access rights.

Yet another solution to get missing comments might be sequential extractions. You can always perform additional extractions after the first one, and latter results will merge with the earlier ones. To do the additional extractions in this specific case:

* Once you have finished extracting the URL

Code: Select all

https://graph.facebook.com/Koran.Fesbuk/albums?since=2012-04-30&until=2012-05-01&limit=100

* Open topic normal under Facebook. It contains a list of Untitled Albums.
* Select all album topics in instances table.
* Start Facebook Graph extractor again. Use old token and select tab URLs.
* Press button Get Subject identifier and Wandora picks up a list of Facebook graph URLs from album topics you selected earlier.
* Change depth to 2 as you want to get the comments of photos too.
* Start extractor by pressing Extract button.

Once the extraction finishes, you should see album topics connected with photos and photos connected with their comments.
Happy experimenting.

Kind Regards,
Aki / Wandora Team