Print

Print


*****  To join INSNA, visit http://www.insna.org  *****

I believe the best way to get at Facebook data is to take a subset,
get consent from people within a Facebook group.  That is, choose a
Facebook group, contact the moderator or creator, and specify what you
are doing to get consent from the users in the group.  In that way,
the members in that group are familiar with what research and methods
you are using.  Forget trying to knock on Facebook to give you data,
that is a 1 in a trillion chance of that happening.  Of course, then
the next challenge is getting this through your local ethics board to
make sure that users know what is happening with their data, that
researchers anonymize the data when used with the FB APIs.

Alvin

On Tue, Feb 10, 2009 at 6:24 PM, Bertil Hatt <[log in to unmask]> wrote:
> *****  To join INSNA, visit http://www.insna.org  *****
>
>> To my mind, if it is crawlable, it should be okay to mine it in
>> the absence of any other Terms of Service prohibition.
>
> Actually, this problem has been faced by search engines for quite some
> time, and a consensus emerged on using a file called "robot.txt". If
> there is one, the intentions of the publishers should be clear there.
> However, most blog platform don't feature one.
>
> Research is not search engines (although most started in a lab)  most
> users won't object to being part of a statiscal result, but if you
> store data, unrecoverable hashes is the recommended pratice. If you
> store the hashing key, you can still identify back sites for a dynamic
> study.
>
>
> On Wed, Jan 28, 2009 at 5:14 PM, Brian Ulicny <[log in to unmask]> wrote:
>> *****  To join INSNA, visit http://www.insna.org  *****
>>
>> Jim wrote:
>>
>> "While any post on the web (such as a facebook or blog entry) is certainly
>> "public" to a pretty high degree, any automated collection of info on lots
>> of individuals raises clear issues with the public; since this is not what
>> the posters expected (fairly or not). The reactions in the press to use of
>> cell-phone logs are a clear case in point."
>>
>> I'd distinguish between a protected site, like Facebook, and an ordinary
>> blog entry on, say, Blogger or Typepad.  Facebook is not crawlable by Web
>> search engine, while Blogger or Typepad blogs are.  Web crawling and
>> indexing is clearly 'automated collection' of data.
>>
>> I don't know what the average blogger perceives about how or whether their
>> blog is crawled and indexed by search engines, and whether this is
>> reasonable or not, but perhaps this is a settled issue legally?
>>
>> To my mind, if it is crawlable, it should be okay to mine it in the
>> absence of any other Terms of Service prohibition.  LinkedIn's Terms of
>> Service includes the following, for example:
>>
>> On the condition that you comply with all your obligations under the
>> Agreement, we grant you a limited, revocable, nonexclusive, nonassignable,
>> nonsublicenseable right to access, through a generally available web
>> browser (but not any scraping, spidering, crawling or other technology or
>> specialist software used to harvest data) to view information that we
>> provide on LinkedIn webpages as we intend such information to be used, and
>> only in accordance with the Agreement and any specific terms of use that
>> we make available to you.
>>
>> Facebook is not crawlable.
>>
>> Brian Ulicny
>>
>>> *****  To join INSNA, visit http://www.insna.org  *****
>>>
>>> Hi Folks -
>>>
>>> Just to weigh in on the privacy issues.  While any post on the web (such
>>> as a facebook or
>>> blog entry) is certainly "public" to a pretty high degree, any automated
>>> collection of info on
>>> lots of individuals raises clear issues with the public; since this is not
>>> what the posters
>>> expected (fairly or not). The reactions in the press to use of cell-phone
>>> logs are a clear case
>>> in point.
>>>
>>> We all have an interest in the long-term viability of social network
>>> research; which depends
>>> crucially on a public that can trust us to use data ethically in a manner
>>> that protects their
>>> privacy.  In that vein; I'd strongly encourage some sort of approach that
>>> gets informed
>>> consent from individuals whenever possible or uses data that is clearly
>>> expected to be a
>>> public trace of social interaction.
>>>
>>> You will have response rate issues, higher cost, slower collection, and
>>> all that.  Which is why
>>> we need good methods research on the effects of such data-collection
>>> biases to our data
>>> collection.  But if a miss-step in data collection turns the public (and
>>> the funding agencies
>>> responsible to the public) against network research, we are all worse off.
>>>
>>> My $0.02.
>>>
>>> All the best,
>>> Jim
>>>
>>>
>>> On 28 Jan 2009 at 11:04, Nicole Ellison wrote:
>>>
>>>> *****  To join INSNA, visit http://www.insna.org  *****
>>>>
>>>> Hi,
>>>>
>>>> Currently Facebook doesn't allow screen-scraping for researchers or
>>>> others. See http://www.facebook.com/terms.php.
>>>>
>>>> In the past they did permit this method of data collection, which is
>>>> why there are some earlier papers out there that use this method (such
>>>> as Lampe et al: https://www.msu.edu/~nellison/lampe_et_al_2007.pdf).
>>>>
>>>> There is one public FB dataset available:
>>>> http://cyber.law.harvard.edu/node/4682
>>>> . However, it's controversial
>>>> (http://fstutzman.com/2008/10/07/facebook-dataset-identified/
>>>> ).
>>>>
>>>> Currently Facebook is extremely cautious about releasing datasets due
>>>> to privacy concerns. One potential way around this constraint is to
>>>> physically go to the FB offices and conduct all analyses on site, but
>>>> this would entail a lot of coordination with the company.
>>>>
>>>> Hope this is helpful, Nicole
>>>>
>>>>
>>>> Nicole Ellison, PhD
>>>> Assistant Professor,
>>>> Department of Telecommunication, Information Studies, and Media
>>>> Michigan State University
>>>> [log in to unmask]
>>>>
>>>>
>>>> On Jan 28, 2009, at 9:24 AM, Terrill L. Frantz wrote:
>>>>
>>>> > *****  To join INSNA, visit http://www.insna.org  *****
>>>> >
>>>> > Brian,
>>>> > The paper featured in the NYT article was published in Social
>>>> > Networks last October.  The authors negotiated with Facebook to get
>>>> > at the data, which certainly is an option for scholars, though is
>>>> > resource costly.
>>>> >
>>>> > Non-special permissioning in Facebook and, consequentially, the
>>>> > api's that follow, is set to limit regular people to identifying
>>>> > their friends and their friends.  The apis are fairly weak, so
>>>> > you'll have to resort to page scraping to get at much of the
>>>> > attribute data for the pages that you can view. Researchers can
>>>> > certainly appeal to Facebook management to remove these security
>>>> > handcuffs.  Otherwise, if you want real-time data without the
>>>> > getting special-permissions, we are left with using some mediocre
>>>> > apis.
>>>> >
>>>> > Cheers,
>>>> > Terrill.
>>>> >
>>>> >
>>>> > [log in to unmask] wrote:
>>>> >
>>>> >> I have been curious about Facebook data mining, too.  There was an
>>>> >> article
>>>> >> in the NY Times a while back profiling various scholars mining
>>>> >> Facebook
>>>> >> data.  Here is the URL:
>>>> >>
>>>> >> http://www.nytimes.com/2007/12/17/style/17facebook.html
>>>> >> On Facebook, Scholars Link Up With Data
>>>> >> By STEPHANIE ROSENBLOOM
>>>> >> Published: December 17, 2007
>>>> >>
>>>> >> The scholars mentioned seem to have access to more than just their
>>>> >> personal friends and friends of friends.  I'm not sure what
>>>> >> Facebook's
>>>> >> terms of service are, nor what their APIs allow.
>>>> >>
>>>> >> If nothing else, you might want to contact the scholars mentioned
>>>> >> in the
>>>> >> article to see how they access their data (or read their papers).
>>>> >>
>>>> >> I hope you will report back on what you find.
>>>> >>
>>>> >> Best,
>>>> >>
>>>> >> Brian Ulicny, PhD
>>>> >> VIStology, Inc.
>>>> >>
>>>> >>
>>>> >>> *****  To join INSNA, visit http://www.insna.org  *****
>>>> >>>
>>>> >>> Hi Ines,
>>>> >>>
>>>> >>> Facebook will only allow you to access your friends and their
>>>> >>> friends.
>>>> >>> ORA (http://www.casos.cs.cmu.edu/projects/ora/) has simple
>>>> >>> Facebook, and
>>>> >>> other real-world source, download features that will automatically
>>>> >>> construct your ego network dataset, then visualize and analyze it.
>>>> >>> Contact me off-line for more info.
>>>> >>>
>>>> >>> Terrill
>>>> >>> Carnegie Mellon University
>>>> >>> http://www.cs.cmu.edu/~terrill/
>>>> >>>
>>>> >>>
>>>> >>> Ines Mergel wrote:
>>>> >>>
>>>> >>>
>>>> >>>> *****  To join INSNA, visit http://www.insna.org  *****
>>>> >>>>
>>>> >>>> Dear Networkers:
>>>> >>>>
>>>> >>>> I
>>>> >>>> am in the process of preparing a short proposal for presentation
>>>> >>>> for
>>>> >>>> our IRB board and was wondering if anyone has insight into
>>>> >>>> whether it
>>>> >>>> is allowed to automatically download Facebook data in the US?
>>>> >>>>
>>>> >>>> It
>>>> >>>> would be great to know a) the legal situation (is it allowed); b)
>>>> >>>> people's experiences with IRB boards, and c) if anyone has used
>>>> >>>> Facebook Connect (or other mechanisms) to download data?
>>>> >>>>
>>>> >>>> Thank you so much for your help! I will make sure to compile
>>>> >>>> responses
>>>> >>>> and send them back to the list.
>>>> >>>>
>>>> >>>> Ines
>>>> >>>>
>>>> >>>>
>>>> >>>> ======
>>>> >>>>
>>>> >>>> Ines Mergel
>>>> >>>> Assistant Professor,
>>>> >>>> Department of Public Administration
>>>> >>>> The Maxwell School of Syracuse University
>>>> >>>> 215 Eggers Hall
>>>> >>>> Syracuse, New York
>>>> >>>> 13244-1090
>>>> >>>> Phone: 315-443-5100 |  Fax: 315-443-9721
>>>> >>>> http://faculty.maxwell.syr.edu/iamergel
>>>> >>>>
>>>> >>>> _____________________________________________________________________
>>>> >>>> SOCNET is a service of INSNA, the professional association for
>>>> >>>> social
>>>> >>>> network researchers (http://www.insna.org). To unsubscribe, send
>>>> >>>> an email message to [log in to unmask] containing the line
>>>> >>>> UNSUBSCRIBE SOCNET in the body of the message.
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>> _____________________________________________________________________
>>>> >>> SOCNET is a service of INSNA, the professional association for
>>>> >>> social
>>>> >>> network researchers (http://www.insna.org). To unsubscribe, send
>>>> >>> an email message to [log in to unmask] containing the line
>>>> >>> UNSUBSCRIBE SOCNET in the body of the message.
>>>> >>>
>>>> >>>
>>>> >>
>>>> >>
>>>> >>
>>>> >
>>>> > _____________________________________________________________________
>>>> > SOCNET is a service of INSNA, the professional association for social
>>>> > network researchers (http://www.insna.org). To unsubscribe, send
>>>> > an email message to [log in to unmask] containing the line
>>>> > UNSUBSCRIBE SOCNET in the body of the message.
>>>> >
>>>>
>>>> _____________________________________________________________________
>>>> SOCNET is a service of INSNA, the professional association for social
>>>> network researchers (http://www.insna.org). To unsubscribe, send
>>>> an email message to [log in to unmask] containing the line
>>>> UNSUBSCRIBE SOCNET in the body of the message.
>>>
>>>
>>> James Moody
>>> Associate Professor of Sociology
>>> Duke University
>>> http://www.soc.duke.edu/~jmoody77/
>>>
>>> _____________________________________________________________________
>>> SOCNET is a service of INSNA, the professional association for social
>>> network researchers (http://www.insna.org). To unsubscribe, send
>>> an email message to [log in to unmask] containing the line
>>> UNSUBSCRIBE SOCNET in the body of the message.
>>>
>>
>> _____________________________________________________________________
>> SOCNET is a service of INSNA, the professional association for social
>> network researchers (http://www.insna.org). To unsubscribe, send
>> an email message to [log in to unmask] containing the line
>> UNSUBSCRIBE SOCNET in the body of the message.
>>
>
> _____________________________________________________________________
> SOCNET is a service of INSNA, the professional association for social
> network researchers (http://www.insna.org). To unsubscribe, send
> an email message to [log in to unmask] containing the line
> UNSUBSCRIBE SOCNET in the body of the message.
>

_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.