Skip to content

the potential of GIS and community information networks

October 31, 2010

I agree with Sasha Costanza-Chock’s response to Malcolm Gladwell: “The revolution will be tweeted – but tweets do not the revolution make.” If they do, then radical participatory democracy is looking a lot different than I expected.

Is bieberspotting really the best we can do with Twitter’s formidable real time search and tracking potential?

MediaPost describes the freshly launched JustSpotted.com as a “real-time search engine turned social site [that]  will identify the whereabouts of celebrities through a relationship with Twitter that gives it access to public tweets. The technology relies on natural-language filters to determine tweets from celebrity sightings.” For example, if I were inclined to stalk Kate Mara, I could easily discover that she was sighted last Thursday at Madison Marvelous in Los Angeles.

I think we can more productively employ mobile reporting and mapping (part of a technology set known as Geographic Information Systems [GIS]) if we think in terms of community information networks. I’m using the term loosely here to mean simply the communicative flows by which people pick up information as they participate in various communities, especially local communities of place. What kinds of information could be usefully tracked and analyzed using reports from mobile devices? What’s missing from out current networks? What could be made more efficient?

Of course, we know that information sharing via social networks helps spread news in crisis situations, like natural disasters and political uprisings. That’s a good thing, but I have something less sensational in mind here. How can this technology be utilized to improve the more mundane aspects of daily existence? What about local communities building shared databases of information? For example, could we track reports of roads and road signage in need of repair? Might that increase the efficiency of municipal works departments? Residents of a community could tweet the location and problem along with a tag. Software could track the information and assess priority – based on the nature of the problem and/or the number of reports.

I’ve seen a few examples of this kind of thinking. A few years ago an online network emerged in which observers shared gas prices at particular locations. Other members of the network could then fill up at the cheaper locations. Ostensibly, this would not only save money for individual members of the network, but it would also increase competitive pressure and force prices lower throughout the network.

According to Time, some Mexicans have been using twitter to report the locations of DUI roadblocks. That’s definitely the spirit I have in mind, if not quite the application. CNN reports that citizens in Reynosa have been reporting dangerous criminal activity – not to the police, but to each other – so that they can avoid trouble. The neighborhood of Miraflores, in Lima, Peru, has been using Alerta Miraflores, an officially integrated crime reporting system licensed by Voxiva Inc., since 2003. They’ve apparently had great success, with over 133,000 reports filed and a 68% drop in robberies  over the first four years. The Voxiva system is quite sophisticated. It uses voice calls instead of text messages and requires a command center, so it’s not as light and decentralized as I have in mind, but it provides a worthy model.

In fact, an organization called CiviRep is (was?) launching a text-based crime reporting system in the Petare, a largely impoverished district of nearly 400,000 residents in Caracas, Venezuela. There’s been no activity on CiviRep’s website since spring of 2009, so this project may be moribund. In any case, here’s are some key excerpts from their project summary:

CiviRep will allow citizens in Petare and eventually in Caracas, Venezuela to report crimes via short-message-service (SMS). This information will then processed and aggregated at the CiviRep engine and displayed on a web-map in real time. The municipal authorities, as well as the local police force will receive this data in real time.  This will elicit more rapid and efficient responses from the authorities in order to serve their constituents.

In the future, we hope to build a feedback mechanism through which the implementing organization can update the status of crime on the CiviRep website thereby providing real time information about the response to a crime. This would enable CiviRep to provide ‘turn-around’ times for each crime to interested parties monitoring the state of affairs in Caracas.

There’s also this informational video:

Of course, there are some major obstacles for such a system. CiviRep themselves identified two of the top obstacles as authentication and anonymity. On the one hand, you need a high percetnage of authentic reports. The Alerta Miraflores system achieved 98 percent authenticity over the first four years but it was not an anonymous system. If the system is not anonymous, however, many potential users will avoid making reports due to fear of retribution. (It’s unclear to me how Alerta Miraflores gained widespread buy-in without anonymity. This is worth looking into.) If the system is anonymous, of course, it’s open to abuse. For example, in Mexico false tweets have been released criminal cartels aiming to incite panic.

Here’s how CiviRep is (was) planning to deal with these issues:

Validation by the police: Our partners in Harvard have received assurance from the local police of Caracas who have agreed to validate the crimes reported by the citizens. Crimes reported by the citizens will be forwarded to the police. The police department can view the crimes on a webpage. They will be provided with a validation user ID and password which they will utilize to validate the crimes. An authentic crime will raise the credit rating of the citizen who has reported the crime while erroneous crimes will decrease the credit rating. Unlike the federal police, the local police is very eager to modernize their systems and find a way to improve and streamline their operations and response times to crime.

Anonymity of citizens: In order to protect the citizens reporting crime from the crime perpetrators, citizen information will be encrypted and saved within the CiviRep database. The encrypted information will provide a mechanism for CiviRep to track credit rating to the citizen while maintaining the anonymity at the same time.

I read recently – though I can’t track down the link – that Mexico, in consultation with the US, was looking to employ similar technology. Applications need not be crime related, however. What else can this technology do? In Africa it’s being employed to prevent medicines shortages. Let’s hope the next example I come across is more akin to that type of endeavor than it is to bieberspotting.

*****

[Editorial addendum: It’s very interesting to see a development project seeking to work with the Venezuelan government make such a candid assessment regarding the difference between federal and municipal police, especially when it reflects so negatively on the feds. Venezuela’s National Police are controlled by the Chavez administration, whereas municipal forces are controlled by local officials, themselves often members opposition parties. This has led to some major confrontations in the past and often comes up in relation to the ongoing debate regarding crime and security. I do not know, however, if the municipal police in Petare are understood to be aligned with either side.]

*****

[UPDATE: Looks like New York City could benefit from a community integrated crime reporting system.)

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: