2. Open Data Providers

Before deciding whether to use any particular open 911 dataset it is important to understand who is responsible for providing that data. This will be important should you need to contact someone with questions about data usage, the data portal itself, or for clarification on a specific data point.

Our Reimagine 911 team grouped the providers into 10 categories in order to compare them at a higher level.

Fig. 2.1 Cities are the most common provider of open 911 data.

Data Providers

The Reimagine 911 team defines the “data provider” as the organization responsible for the uploading of the data and/or maintaining the open data portal. This is often different from the organization that generates or owns the data.

There is ambiguity regarding which organization owns, is in charge of, or is responsible for the open data. In our Open Data Review, we found that open data sets often had two or three entities associated with each dataset:

1. The department that actually collects the data

2. The department that is responsible for uploading the data and/or the open data portal

3. An individual who may have done the initial uploading of the data

Sometimes the department that collects the data was also responsible for uploading the data and sometimes they are not.

An Explicit Provider

In Fig. 2.2 below from the data portal for Scottsdale, AZ we see a clearly stated provider.

Fig. 2.2 Demonstrating a provider that is different from the agency that collected the data.

The data is collected by the Fire Department, but according to their open data portal, it is owned by the City GIS department. In the Reimagine 911 review of cities, we would list the “City of Scottsdale GIS” as the provider because we decided to identify organizations, not individuals.

Our research found a variety of terms indicating ownership of the data such as “Publisher,” “Dataset Owner,” “Member,” or “Private Member.” Language often varies across open data platforms.

An Ambiguous Provider

In many cases the provider of the data is not as clear. Another example (Fig. 2.3) is from the Honolulu Police Department's Crime Incidents dataset.

Fig. 2.3 Demonstrating a common case where it is difficult to determine the dataset provider.

Here, the “owner” is an individual and no agency is listed. Since the Reimagine 911 program prioritized identifying the entity responsible for uploading the data or maintaining the data portal, our program used the Honolulu Police Department as the provider.

Distribution of Agencies

When working with open 911 data, you will likely find that the data provided represents multiple agencies; very rarely is a dataset exclusive to 911 call data. Although police departments provide the majority of open 911 data, it is not uncommon to find data provided by fire and medical departments as well. There is a variety of combinations: datasets that cover one, two, and all three agencies.

Fig. 2.4 Our review found that about half the datasets contained data from multiple agencies.

There are typically two ways that multiple agencies are represented in open 911 data:

1. The open data provider combines data from multiple agencies into a single dataset

2. The open data provider makes multiple datasets available from the different agencies

There may be data overlap across different datasets in a city due to other departments being asked to assist, etc. While 35 cities provided more than one dataset, it is highly likely that they contain data on the same incidents. See Appendix A for a table of helpful search terms for identifying datasets from specific agencies.

Fig. 2.5 Police is the most common agency represented in the datasets.

Identifying the Data Provider

From the 127 data providers we initially reviewed, we were unable to find providers for four of them. They were generally from third-party providers.

When checking for a provider, look for an “about” section within the open data portal, a page on the city’s website about their open data program, or the policies section of the city’s website—it might have details about the use and maintenance of open data.

When we reviewed the open datasets to determine whether they contained data from police, fire, EMS (or all three combined), it was helpful to have a short list of search terms. When the following terms are present in the data, they tend to be good indicators that the associated agency is represented. Our actual determination of the agencies represented depended on both the number and prevalence of these terms and others not listed here.

Agency	Commonly Used Search Terms
Police	welfare check suspicious behavior juvenile	law assault PD
Fire	smoke	fire
Medical / EMS	breath heart mental suicide ambulance	medic ems burn injured sick