Centrality Data Science is creating Open city council voting records
8
patrons
$81
per creation
In 2014, we launched CityBlocs, a website to help citizens keep track of how their city councils vote through visualizations of their voting patterns.

In most cities, tracking how city councillors vote is actually a pretty time consuming process. Even though cities provide minutes of the meetings online, they are almost always presented as unstructured data, meaning that to get the councillors' voting records, you have to go through them by hand and translate the text into usable data. This makes getting this information laborious and error-prone.

Now, we want to expand the project: we want to unlock voting records from city council minutes using tools we've built, develop topic classification algorithms to classify what the votes are about, build other tools to make the data more useful and make all of the data freely available as easy-to-use data sets.

So far, support from patrons has allowed to create and maintain a dataset of council voting records for London, Ontario. We're also holding a vote to add a second city in February 2016!

The data will be regularly updated and posted for all to use at cityblocs.ca. At the end of every month in which we reach a milestone funding goal, all CityBlocs patrons can vote to decide which cities or feature upgrades will be added next.

Currently, the datasets contain:
  1. every motion that comes before council;
  2. a classification of the motion by topic using automated methods developed by Centrality;
  3. how each councillor voted.
The dataset for the first city funded under this project is available here: http://cityblocs.ca/api/v0.1/. If there is something else that would make the dataset more useful to your project, please let us know!

There are many different types of questions that these data can more easily provide the answer to than going through minutes by hand, or -- even worse! -- attending every council meeting in person. As the project develops, it will consist of more than just datasets, and we have plans to add features to the website, such as:

  1. REST API: This will allow users to download subsets of the database in a variety of formats, in addition to downloading the CSV files.
  2. Browsable interface: This will allow users to directly access the data from the website without using the API.
  3. More advanced query options: For instance, you could query the database to see what percentage of votes were lost by the sitting mayor or how often pairs of councillors vote together. (You can do this with the CSV files if you want, but this will make it easier.)
  4. Upgraded vote classifier: Automatically classifying text is a difficult problem, and adding additional power to the vote classifier will make the classifications even more accurate.
  5. Councillor profile pages: This feature will summarize city councillors at a glance. Ever wanted to know which councillors were most similar in their voting patterns? Which councillors end up voting on the winning side most often? This tool will help.
Using these data, we hope citizen groups and researchers will make tools and analyze the data to help us all better understand our governments, journalists can more easily see what's going on their city councils and all of us can benefit from knowing what's going on in city hall.

We'd love to continue to do all of this with your help.

What's involved in the project?

We've been hard at work developing a Python library useful for extracting data from city council minutes. The project so far has also involved writing a simple motion classification system using a dictionary method, classifying motions by keyword. We also need to maintain the database infrastructure, and expand the website to include a number of useful tools. Funding will allow us to cover these ongoing costs, and all money raised goes directly into making CityBlocs better and available for more cities.

Who's working on the project?

This is a project by Centrality Data Science, based in Calgary. The team consists of Paul Fairie (PhD, Political Science; twitter: @paulisci), Adam D'Souza (PhD, Physics; twitter: @h2o_bro) and Michael Underwood (PhD, Physics; twitter: @munderwood).
Tiers
Pledge $1 or more per creation
VOTE!

For every month you contribute at this level, you get one vote. Every time we reach a new goal level, all contributors get to vote on what city gets added to the project.
Pledge $4 or more per creation
10 VOTES!

For every month you contribute at this level, you earn 10 votes to cast once a new milestone is reached.
Pledge $7 or more per creation
20 VOTES!

For every month you contribute at this level, you get 20 votes.
Pledge $10 or more per creation
30 VOTES!

For every month you contribute at this level, you get 30 votes.
Pledge $20 or more per creation
EARLY ACCESS

For every month you contribute at this level, you also get early access for 2 days to one city covered by the CityBlocs project, so you get a chance to analyze the data before everybody else. You also get 60 votes.
Pledge $25 or more per creation
EARLY ACCESS TO ALL CITIES

For every month you contribute at this level, you also get early access for 2 business days to all cities covered by the CityBlocs project, so you get a chance to analyze the data before everybody else. You also get 75 votes.
Pledge $300 or more per creation
only 1 left
NAME A CITY

For every month you contribute at this level, you also get exclusive early access for 7 business days to a city of your choice (the earliest anyone else will receive the data is 5 days after you, for those patrons at either the $20 or $25 level), so you get a chance to analyze the data before everybody else. You also get 100 votes.
Goals
$81 of $90 per creation
Patrons will vote to decide the first city to be launched along with a motion classification system and website. For each funded city, a CSV file with every vote will be posted indicating the motion, its text, how each councillor voted, and a topic classification of each motion. The cities will be updated at least once a month for the duration of the project (depending on the council's meeting schedule).
1 of 5
In 2014, we launched CityBlocs, a website to help citizens keep track of how their city councils vote through visualizations of their voting patterns.

In most cities, tracking how city councillors vote is actually a pretty time consuming process. Even though cities provide minutes of the meetings online, they are almost always presented as unstructured data, meaning that to get the councillors' voting records, you have to go through them by hand and translate the text into usable data. This makes getting this information laborious and error-prone.

Now, we want to expand the project: we want to unlock voting records from city council minutes using tools we've built, develop topic classification algorithms to classify what the votes are about, build other tools to make the data more useful and make all of the data freely available as easy-to-use data sets.

So far, support from patrons has allowed to create and maintain a dataset of council voting records for London, Ontario. We're also holding a vote to add a second city in February 2016!

The data will be regularly updated and posted for all to use at cityblocs.ca. At the end of every month in which we reach a milestone funding goal, all CityBlocs patrons can vote to decide which cities or feature upgrades will be added next.

Currently, the datasets contain:
  1. every motion that comes before council;
  2. a classification of the motion by topic using automated methods developed by Centrality;
  3. how each councillor voted.
The dataset for the first city funded under this project is available here: http://cityblocs.ca/api/v0.1/. If there is something else that would make the dataset more useful to your project, please let us know!

There are many different types of questions that these data can more easily provide the answer to than going through minutes by hand, or -- even worse! -- attending every council meeting in person. As the project develops, it will consist of more than just datasets, and we have plans to add features to the website, such as:

  1. REST API: This will allow users to download subsets of the database in a variety of formats, in addition to downloading the CSV files.
  2. Browsable interface: This will allow users to directly access the data from the website without using the API.
  3. More advanced query options: For instance, you could query the database to see what percentage of votes were lost by the sitting mayor or how often pairs of councillors vote together. (You can do this with the CSV files if you want, but this will make it easier.)
  4. Upgraded vote classifier: Automatically classifying text is a difficult problem, and adding additional power to the vote classifier will make the classifications even more accurate.
  5. Councillor profile pages: This feature will summarize city councillors at a glance. Ever wanted to know which councillors were most similar in their voting patterns? Which councillors end up voting on the winning side most often? This tool will help.
Using these data, we hope citizen groups and researchers will make tools and analyze the data to help us all better understand our governments, journalists can more easily see what's going on their city councils and all of us can benefit from knowing what's going on in city hall.

We'd love to continue to do all of this with your help.

What's involved in the project?

We've been hard at work developing a Python library useful for extracting data from city council minutes. The project so far has also involved writing a simple motion classification system using a dictionary method, classifying motions by keyword. We also need to maintain the database infrastructure, and expand the website to include a number of useful tools. Funding will allow us to cover these ongoing costs, and all money raised goes directly into making CityBlocs better and available for more cities.

Who's working on the project?

This is a project by Centrality Data Science, based in Calgary. The team consists of Paul Fairie (PhD, Political Science; twitter: @paulisci), Adam D'Souza (PhD, Physics; twitter: @h2o_bro) and Michael Underwood (PhD, Physics; twitter: @munderwood).

Recent posts by Centrality Data Science

Tiers
Pledge $1 or more per creation
VOTE!

For every month you contribute at this level, you get one vote. Every time we reach a new goal level, all contributors get to vote on what city gets added to the project.
Pledge $4 or more per creation
10 VOTES!

For every month you contribute at this level, you earn 10 votes to cast once a new milestone is reached.
Pledge $7 or more per creation
20 VOTES!

For every month you contribute at this level, you get 20 votes.
Pledge $10 or more per creation
30 VOTES!

For every month you contribute at this level, you get 30 votes.
Pledge $20 or more per creation
EARLY ACCESS

For every month you contribute at this level, you also get early access for 2 days to one city covered by the CityBlocs project, so you get a chance to analyze the data before everybody else. You also get 60 votes.
Pledge $25 or more per creation
EARLY ACCESS TO ALL CITIES

For every month you contribute at this level, you also get early access for 2 business days to all cities covered by the CityBlocs project, so you get a chance to analyze the data before everybody else. You also get 75 votes.
Pledge $300 or more per creation
only 1 left
NAME A CITY

For every month you contribute at this level, you also get exclusive early access for 7 business days to a city of your choice (the earliest anyone else will receive the data is 5 days after you, for those patrons at either the $20 or $25 level), so you get a chance to analyze the data before everybody else. You also get 100 votes.