You are viewing a read-only archive of the Blogs.Harvard network. Learn more. Facilitating Information Seeking For Hyperlocal Communities Using Social Media

April 15th, 2013 by andresmh

You hear sirens blaring in your neighborhood and, naturally, you are curious about the cause of commotion. Your first reaction might be to turn on the local TV news or go online and check the local newspaper. Unfortunately, unless the issue is of significant importance, your initial search of these media will be probably be fruitless. But, if you turn to social media, you are likely to find other neighbors reporting relevant information, giving firsthand accounts, or, at the very least, wondering what is going on as well.



Social media allows people to quickly spread information and, in urban environments, its presence is ubiquitous. However, social media is also noisy, chaotic, and hard to understand for those unfamiliar with, for example, the intricacies of hashtags and social media lingo. It should be no surprise that, regardless of the popularity of social media, people are still using TV and newspapers as their main sources for local information, while social media is just beginning to emerge as a useful information source.  We created to address this issue. reveals the latent neighborhood-specific information that already exists in social media. In our first prototype of the tool, we focused on Twitter posts to derive relevant news, people, and events that are set within a particular locality, often referred to as “hyperlocal information”.

Inspired by the typical journalistic questions, i.e., “what, who, where, and when” uses various machine learning algorithms to provide four types of hyperlocal content in a simple web-based interface:

  • Active events: events that are trending in the locality.
  • Top topics: most frequently mentioned terms and phrases from recent Twitter posts.
  • Popular placesmost frequently checked-in/mentioned.
  • Active people: Twitter users mentioned the most).

We investigated the effectiveness of as a tool for finding neighborhood information through a user study of 13 residents from three Seattle neighborhoods: Capitol Hill, Wallingford, and Rainey Valley. Participants found to perform better than Twitter at these particular four tasks related to neighborhood information:

  • finding recent events,
  • finding local neighborhood reporters,
  • finding neighborhood topics, and
  • finding potential neighborhood friends.

The overall reaction to the information provided on was quite positive.. The participants in our study found easier to use than Twitter and the majority said they would prefer it as a tool for exploring their neighborhoods.

One of the interviewees mentioned: was set up specifically with the community in mind. It makes community news/events/issues/people etc. easily accessible.

Participants generally found it easier to complete
neighborhood exploration tasks using was found to be more useful, easy to use,
with a better overview of the users’ neighborhoods, and a
sense of connection to their neighborhood communities.

Tools like are beginning to uncover the future of computational civic media that we believe will be an important component of the information ecosystem.

For more, see our full paper Facilitating Information Seeking For Hyperlocal Communities Using Social Media to be presented at CHI 2013.

Yuheng Hu, Arizona State University
Shelly D. Farnham, Microsoft Research
Andrés Monroy-Hernández, Microsoft Research

Crossposted from the Follow the Crowd Blog

Comments are closed.