New crowd sourcing model launched for Boston 311

Mayor Martin J. Walsh and the City's Department of Innovation and Technology (DoIT) today launched a new crowdsourcing effort to gather the data that will power the City's new 311 web interface. The new interface, developed in partnership with BOS:311 and the City's Department of Innovation and Technology, will use a computer model that takes a description of the user's issue and suggests case types that are most likely to fit that description.

 

"We're committed to using data to improve City services and make the lives of our residents easier," said Mayor Walsh. "The stronger dataset we have to power the 311 system, the more efficient the service will be."

 

This new interface will change the way each resident will interact with the 311 program. The current system prompts the user to choose from dozens of the most frequent 311 case types and find the one that matches their issue, in order to direct them to the right form.

The new system will have the user tell 311 in their own words what's wrong, in a sentence or less, and a computer model will suggest the correct case type. Users will no longer need to scroll through the full set of case types to find the one they are looking for, making it easier and more efficient to submit requests.

 

"We're excited for the opportunity the new model brings to the 311 service," said Jerome Smith, the City of Boston's Chief of Civic Engagement. "We're always looking to provide constituents with helpful and responsive interactions and I look forward to seeing how this new feature can improve the user-experience." 

 

In order for the new 311 model to predict what case type the user is searching for, the model needs data. This data will link certain keywords or phrases to case types. For example, the phrase "no one has cleaned the snow off my sidewalk" would most likely indicate a "Sidewalk not shoveled" case, so users would see that as the top suggestion in the list.

As the new 311 model gets more data, it will learn what words and phrases are most common when talking about different case types. Over time, the model will even be able to recognize patterns. If the model notices that snow-related cases typically come in batches (like during and after a storm), those suggestions will more likely be shown when a spike in snow cases comes in. 

 

"The new 311 model is unique in that it will translate the perspective of the constituent into the language of the 311 system" said Dan O'Brien, co-director of the Boston Area Research Initiative (BARI) and faculty in the School of Public Policy and Urban Affairs at Northeastern University. "Using data to connect these perspectives makes public maintenance even more efficient."

 

To gather the data needed for the new 311 system, the City has released a web application that provides a user with a case type and asks for three different ways the user may describe it.

 

The City plans to release the dataset on Analyze Boston, the City's open data hub, so that other cities can use it for their own systems in the future, and the model code will also be released as open source software once completed.

 

BOS:311 is the go-to place for constituents to report non-emergency services. The service enables real-time collaboration with citizens as they report potholes, graffiti, and other issues from anywhere in the city. Constituents can file a report using the 311 website, 311 mobile application, or by tweeting to@BOS311.

 

The Department of Innovation and Technology makes sure the networks, computers and systems that support the City are secure and effective. They also manage the City's websites and technologies focused on service delivery.

 

The department's core work involves five priority areas: