Monitoring Social Media by Location–Do tools like GeoFeedia invade privacy?

Post by: Kim Stephens

I recently saw a presentation about GeoFeedia,  a social media monitoring software system, at an SMEM-type event sponsored by the Metropolitan College of NY’s Emergency and Disaster Management MPA program. Their presentation, along with the ensuing conversation, inspired me to take a closer look.  GeoFeedia’s monitoring tool is unique among all of its competitors: it allows the user to include location searches for content versus only searching by keyword. Timo Luge’s blog post about the service describes why this feature is so important:

If someone uploads and photo a writes “Here is a photo of my house after the #earthquake in #Alphaville” it’ll be easy to find, but if someone simply writes “My house – so sad!“ you won’t find it using standard tools. And while you can use services like Hootsuite to show you all tweets in certain area, doing this is quite tricky and too complicated for most users.

Watch their video for a great one-minute description.

I encourage you to explore their website for details about what the tool can do, but below I briefly describe its 6 main features, including location monitoring:

1. It enables “advanced location monitoring” which will gather–as well as archive—social communications based on the location determined by the user. For instance, if there was an explosion in an area, law enforcement could hone in on people posting about the incident who are in close proximity.  In contrast, if only a keyword search were used, then the collected content would include everyone discussing the event.
2.  Keyword searches are not eliminated from GeoFeedia. The ability to filter social content and refine searches using keywords as well as timeframe, media type, author “and more” is included.
3. Data is exportable in a feature they call “data portability.”
4. Content from GeoFeedia can easily be published to your own social networks.
5. Analytical tools are embedded that can help the user identify such things as the most active or influential poster, as well as trends, etc.
6. The content of the geo-feed is shareable to people “inside or outside your organization. You can even share a live Geofeed stream with users that don’t have a Geofeedia account.” In contrast, other software tools, including Radian 6, have very restrictive sharing policies and explicitly do NOT allow sharing of content with non-account holders.

Do Geo-location Tools Invade Privacy?

Following their presentation in New York,  the question of privacy was raised, which I thought was interesting.  Do people expect some privacy when posting content to social networks?  If you are quick to answer “no” then consider this statement by one of the participants: “Ask any teenager if posting to Twitter and Instagram allows everyone to see their content and they will answer ‘yes.’ Then ask them if they would be OK with their parents viewing whatever they are posting and they will emphatically answer ‘NO!'” I know this to be a fact since I happen to live with two teenagers.  

Nonetheless, during a crisis situation it would seem that folks would desire government officials to see their posts in order to receive assistance. But in the world of law enforcement, this type of monitoring tool takes on an entirely different connotation. I can image that a law enforcement agency could use GeoFeedia, or something like it, to draw a virtual circle around a park where a festival is happening and look for information about illegal activity. Would people posting images and Tweeting at the event, for instance, realize they were being “watched” virtually? 

 Based on current law, it seems that these types of searches are legal and possibly even expected by citizens.  Law professor Jonathon Turely wrote a couple of years ago about our evolving societal privacy expectations in a Washington Post article about the then upcoming Supreme Court case Jones v. United States. The case involved a conviction of man based on his activities that were tracked with a GPS devise:

“This surveillance continued after a warrant had expired. But the Obama administration insists that no warrant should be required for the government to track the movements of citizens with such devices. The administration says that the new technology merely captures what can be observed, albeit in far greater detail. But the technology could allow the government to follow an almost limitless number of citizens in real time, all the time.” 

Professor Turely asks:  “As we come to expect less privacy, are we entitled to less of it?” The court did eventually hold that “the Government’s installation of a GPS device on a target’s vehicle, and its use of that device to monitor the vehicle’s movements, constitutes a ‘search'” under the Fourth Amendment.” However, the broader privacy issue was left unsettled.

How does that case relate to social media? Stephen E. Henderson’s in his journal article Expectations of Privacy in Social Media,  cites the Jones case and many others in his argument that the public really can’t expect privacy when posting publicly on social networks, however, they probably can expect privacy when an effort is made to keep the communication out of the public view, for instance with  protected Tweets and Facebook messages. The notion of control of the information is the key. “If the government  obtains information that was previously in one’s exclusive control, then it has violated the person’s rights.” However, if the information is publicly available, then there is no interference with a “possessory interest.”  This passage sums up his argument:

Whatever the precise definitions of search and seizure, the Court has articulated this general principle:

[T]he Fourth Amendment protects people, not places. What a person knowingly exposes to the public, even in his own home or office, is not a subject of Fourth Amendment protection. But what he seeks to preserve as private, even in an area accessible to the public, may be constitutionally protected.

This limitation makes eminent sense, in that police should not have to be the only ones to avert their eyes. If you tape a message on a window visible from the street, or place a pie to cool or a plant to grow there, a police officer driving or walking by is free to give it a look. According to the control theory of information privacy, you have chosen to share that information. Whereas if you carry any of those items on your person in public, but in an opaque container, the item remains private, and police must act accordingly.

Some social media is exposed to the public, such as an open-to-the world blog. It is not reasonable to expect privacy when one publishes something to all comers. So there would be no Fourth Amendment restraint on police obtaining the content of such a blog, either by bringing up the site themselves or via the third party hosting that content. The same holds true for a Facebook wall which the user leaves open to the public, YouTube videos left open to the public, and flickr pictures left open to the public. And the same holds true for tweets from a public account, meaning one for which the user does not restrict followers. Since any private person can obtain these things without restraint, the police can as well. (page 238)

Do you have an expectation of privacy? Does your law enforcement agency worry about this topic?  My treatment of this issue is admittedly very superficial–there are many different facets to explore. For a much lengthier treatment of the topic, I encourage you to read Mr. Henderson’s entire article, you can also see sites such as the  Electronic Privacy Information Center.

For another review and more information about GeoFeedia find a comprehensive post by Patrick Meier on iRevolution here. Despite the debate over privacy, I think GeoFeedia is a great tool and probably represents the future of how emergency response organizations will monitor social content in order to gain the best situational awareness. Let me know what you think.

Related articles

• A lesson from Superstorm Sandy: How to find sources using social media (onlinejournalismblog.com)

#SMEM Challenge for 2013: Strategically Monitoring Social Media

Post by: Kim  Stephens

One of the biggest #SMEM challenges for emergency management and public safety organizations is determining whether or not, and increasingly how, they will monitor social media. In the past year we saw a change in mindset: a desire to actively listen versus simply push content to the public. Yet, monitoring can seem like a daunting task.  During large-scale emergency events millions of new posts, pictures and videos  are added to YouTube, Twitter, Facebook, etc. every day. How can a small local public health, first response, or emergency management agency sort through all of that? In this post I will outline strategies I have seen employed to effectively deal with this seemingly insurmountable task.

Where do we start?

Numerous questions come to mind when considering monitoring social networks:

• How ? What software do I need, if any? (And relatedly, how much will that cost?)
• Who? Who will be doing this work?  Will this only be done when the  EOC is stood up and resources can be shared by all response agencies? or  Will resources be required to monitor social media all the time if we have social accounts?
•  What is done with the information gleaned from monitoring social media? How is the information shared (if at all) with response partners when there isn’t an EOC or JIC?

Establishing objectives

Each of the questions above are dependent on the objective(s) established by your organization.  The first step is to determine why: Why are we listening? What is the expected outcome? Only then can you  decide what resources you will need, how you will share that information with response partners, and what tools are required.

Your emergency management organization might decide to specify a number of objectives. Some of the more common ones include:

1.  To determine if the organization’s message is getting across or if conflicting information (rumors) is being conveyed: Are people  confused about what to do (e.g. how long to boil water)?
2. To determine public sentiment regarding the organization or, during a crisis, about the overall government’s response effort: Are people angry about something that is happening?
3. To determine the most commonly asked questions and concerns.
4. To quickly answer direct questions, or questions directed at the community political leadership on topics that involve your organization: Are people asking the Twitter-happy Mayor when debris will be picked up in their neighborhood?
5. To determine what other organizations are saying, in order to both ensure messages are coordinated, and to amplify mission related content.
6. To determine the extent of damage and impact of the disaster event. (Advanced)

It should be noted that law enforcement officials might have completely different set of objectives. They might monitor social media to actively look for people (or evidence) from those who have been involved in a crime as well as to enlist their followers in helping them identify suspects. They could also monitor the accounts of a person that has been brought to their attention by members of the community (e.g. a person has been posting strange comments that point to criminal intentions). In this post, however, I will stick to emergency management concepts since that is much more familiar territory for me.

Low Budget Solutions

Of course, part of the strategy for listening or monitoring social media has to include determining who will be responsible for doing these tasks. I recommend you also read the post that describes VOST (Virtual Operations Support Teams) for some ideas on how you can expand your efforts when required. Nonetheless, there are many things that can be done by an organization to make monitoring social media a bit easier, especially if some of it is completed before a crisis.

The following simple steps are based on processes described by emergency managers who have made the most of the free tools at their disposal. Even though these items might seem like obvious courses of action, I have cited them here for a reason.    I have included some basic 101-type info since people often ask these questions.

1. Create Lists and Like Pages of Response Partners: It is important to know and keep track of what other response organizations are saying on social networks, even if (maybe especially if) they are in a neighboring county. If you and your neighbor put out conflicting content, believe me, the public will notice. (This happens in quickly moving events–road closures are a prime example.)

• On Twitter, set up a list(s) of all “trusted sources” including government agencies, first responders, political leaders,  volunteer organizations and local news media–don’t forget to include federal agencies such as FEMA, EPA and HUD.  Twitter.com explains how to create a list in 4 simple steps.
• Include social streams of all response partners on your website or Facebook page, so the public can easily find them as well. See a best practice example from Australia: Queensland Police Service Alert, which has the embedded Twitter feeds from their response partners organized by sector: transportation, power and water, etc.
• On Facebook, “Like” all of these same organizations.

Coordinate Offline: It should be noted that in addition to doing the work online, a good practice is to have every government official responsible for posting to social networks  participate in recurring meetings to talk about strategies and coordination before a disaster event. (How can we ensure information is updated on our social media accounts, simultaneously? How can we share content/intel that we are seeing from the public ?) The speed of social media might require new, or at a minimum, faster coordination processes. 

2. Invest in a smart phone for the person monitoring social media: Smart phones are a great way to monitor your social media presence when you are away from a computer. Both Twitter and Facebook can provide smart phone notifications to the administrator every time the account is mentioned, replied to, re-tweeted, etc. You can also set up a way to receive notifications when other organizations post updates as well.

• Twitter.com has a great help page on this topic.
• Facebook has similarly helpful “How-To” page about how to receive push notifications on a mobile device. There is also information here about the Page Manager App that lets admins check on their Page activity, view insights and respond to their audience from their mobile device. This app is only currently available for iPhone and iPad.

3. Read

I once asked the social media manager for the US Environmental Protection Agency how he monitored the agency’s social stream, he simply stated: I read.  Surprisingly, keeping up with what is happening on social media does not necessarily take complicated software, especially if reading is done strategically.   In order to prevent being overwhelmed,  you can limit the content that you look at to some or all of the following:

•  Read comments and questions directed to your organization. This step is probably the most important: if your organization is actively posting content, more than likely, people will be posting comments and questions…AND they will expect a response.  Reading comments and “@” messages will also allow you to gauge how your efforts are being received.
• Read what is being posted by your trusted-sources on the list(s) you have created.
• Read comments and questions posed from the public to your response partners and elected officials.
• Read information based on keyword searches and hashtags.   This strategy involves searching for key words, such as the name of the event, in order to find pertinent content.
  • During an active event, people often post pictures and video to Twitter (more so than other platforms) and mention the location and /or name of the town. (For specific instructions see Twitter advanced search and the “How-To“).  It is important to note, however, that any early pictures should be treated cautiously. Some folks think it is quite funny to post fake images.
  • Possible search terms: name of agency, name of event, name of municipality.

4. Actively ask for information

There is nothing wrong with asking your followers or the general public for information via your social networks. People often provide valuable situational awareness information to you anyway, for example, posting on your Facebook page: “There are power lines down on Elk Road.” Some organizations have tried to give the public a way to provide information in a more structured way. Good examples of this are the not-so-new USGS’s earthquake detection program  “Did you Feel it?” and the recent Fairfax County Hurricane Sandy Crowdmap that allowed people to post their observations.

Soliciting information is almost the opposite of  “data mining.” Data mining involves  automated computer processes  intended to make sense of or find patterns in vast amounts of content posted to social networks (see this post by Patrick Meier for more info). I suspect that this process will be one of the hottest topics for 2013 as more of these tools (discussed in this previous post)  come online. Nevertheless, if your organization is simply trying to keep up with mentions and comments, then advanced software is probably not necessary…although highly coveted. Coordination and collaboration with your response partners,  however, continue to be some of the best tools in your toolbox.

If you are still reading, let me know if you have established objectives or listening strategies.

Note: A majority of this content came from a post I did for WRHSAC.org.