Version 1.0

Data Exploration and Integration Lab (DEIL)
Centre for Special Business Projects (CSBP)

October 2, 2020

Table of Contents

1. Overview
2. Data Sources
3. Reference Period
4. Target Population
5. Compilation Methodology
   • Data Standardization and Cleaning
     • Address Parsing
     • Removal of Duplicates
     • Identification of Invalid Entries
     • Other Data Cleaning Steps
     • Selection of Record to Retain in Case of Duplicates
   • Classification Used and Assignment of Cultural and Art Facility Type
   • Geocoding and Determination of Census Subdivision
   • Inclusion in the ODCAF of Facility Type Provided in Source Datasets
6. Database Coverage
7. Data Quality
8. Data Dictionary
9. Contact Us

1. Overview

This experimental Open Database of Cultural and Art Facilities (ODCAF) is one of a number of datasets being created as part of the Linkable Open Data Environment (LODE). The LODE is an exploratory initiative of the Data Exploration and Integration Lab (DEIL) at Statistics Canada. It aims at enhancing the use, accessibility and harmonization of open data from authoritative sources by providing a collection of datasets released under a single licence, as well as open-source code to link these datasets together. This initiative is also meant to explore open data for official statistics and to support geospatial research across various domains. The LODE datasets and code are available through the Statistics Canada website and can be found at: Linkable Open Data Environment.

The ODCAF is a database of cultural and art facilities released as open data. Data sources include various levels of government within Canada^{Footnote 1} and professional associations. This document details the process of collecting, compiling, and standardizing the individual datasets of cultural and art facilities that were used to create the ODCAF. The ODCAF is made available under the Open Government Licence – Canada.

In its current version (Version 1.0), the ODCAF contains approximately 8,000 individual records. The database is expected to be updated periodically as new open datasets become available. The ODCAF is provided as a compressed comma separated values (CSV) file.

2. Data Sources

Multiple data sources were used to create the ODCAF. The sources used are detailed in a 'Data Sources' CSV file located within the zipped data folder available for download on the ODCAF webpage. The links to the original datasets, licenses or terms of use, attribution statements and additional notes are also included in the Data Sources CSV file. For further information on the individual licences, users should consult directly the information provided on the open data portals of the various data providers. In addition to openly licensed databases, the ODCAF also includes a publicly available listing of cultural and art facilities.

The distinction between open and other publicly available data is based on the licensing terms (explicit or implicit) attached to each source dataset used. Open data licenses permit, in varying degrees, usability for any lawful purpose, redistribution (re-sharing) and modification and re-packaging of the data. However, open data licenses can impose some restrictions, such as attribution of original source, share-alike (re-sharing only with like conditions), and no commercial use. Examples of open data licenses are Creative Commons, MIT, GPLv3, and Canada's Open Government License. In general, no warranty is expressed and there are very minor conditions stipulated by the provider.

Publicly available data that are not open data might be associated with proprietary licensing or terms of use that may restrict some of the aspects that would otherwise be permitted under open data licensing.

3. Reference Period

The Data Sources CSV provides, when this is known, either the update frequency or the date each underlying dataset was last updated by the provider (this information is collected at the time the dataset was accessed for this project). Additionally, the Data Sources CSV provides the date each dataset used in the ODCAF was downloaded or provided by the organization that is the source of the data. Data were gathered between January 2020 and July 2020. Users are cautioned that the download date should not be used as an indication of the reference date of the data. To obtain specific information concerning the reference dates of the source datasets, users might contact the relevant data providers directly.

4. Target Population

For the purposes of the ODCAF database, cultural and art facilities are facilities wherein the primary activity is of a cultural nature or is related to the arts. The target population includes only brick and mortar cultural and art facilities that offer programs or services to the general public.

In terms of the North American Industry Classification System (NAICS), the facilities in the ODCAF are primarily in the following sub-sectors:

• 711 - Performing arts, spectator sports and related industries
• 712 - Heritage institutions

Facilities are included when their primary activities have a cultural or arts character, regardless of the source of funding, private or public status, operator type, location or other attributes. However, facilities that are not open to the general public and those that are primarily commercial in nature are not included. Thus, a theatre that offered ballet performances would be in scope, while a ballet school that offered training and performances only to paying students would not.

5. Compilation Methodology

This section provides an overview of the processing done to compile the ODCAF.

Data Standardization and Cleaning

The first processing component for compiling the ODCAF database comprised reformatting the source data to CSV format and mapping the original dataset attributes to standard variable (field) names. This was done using a version of the custom OpenTabulate software developed by the LODE team. A data dictionary of the variables used is provided in section 8.

Owing to the different classification systems and data attributes used in the source datasets and the need to standardize through application of several processing steps, the potential exists for the introduction of errors.

The methodology and limitations of the techniques used in each step used in the data cleaning process are described below. Trivial cleaning techniques, such as removal of whitespace characters and punctuation removal, are omitted from discussion.

Address Parsing

The libpostal address parser, an open source natural language processing solution to parsing addresses, was used to split concatenated address strings into strings corresponding to address variables, such as street name and street number. Occasionally, addresses were split incorrectly due to unconventional formatting of the original address. While effort was made to identify and correct these entries in the final database, some incorrectly parsed entries may have remained undetected. Exceptions are entries with street numbers of the form of two numbers separated by a hyphen or space. Entries of this form usually indicate that the address parser incorrectly parsed a numbered street name (e.g., "123 100 ave" is parsed into the street number "123 100" and the street name "ave", or else that a unit has not been identified correctly (as in "3-100 main st"). Numbers of this form are automatically separated, where the right most number is prepended to the street name if the street name is a variant of the word "street" or "avenue." Otherwise, the left most number is appended to the unit column.

A limited number of entries were manually edited when it was clear that the parsing had not been done correctly. An example is addresses with hyphenated numbers such as "1035-55 street nw", which may have been interpreted as having a civic number of "1035-55" and a street name of "street nw", rather than a civic number of 1035, and a street name of "55 street nw". While effort was made to ensure that the results are correct, it is possible that the scripts used to process and parse the addresses may unintentionally cause other, undetected, errors. Should any such errors be reported to or detected by the LODE team subsequently, they will be corrected in future versions of the ODCAF.

Removal of Duplicates

The removal of duplicates was done using both literal and fuzzy string matching on the facility name and street name, conditioned on the street number and province; by "conditioned," it is meant that a fuzzy comparison between two facilities is made provided that the street numbers and provinces agree. The fuzzy comparison is done using the Python package FuzzyWuzzy^{Footnote 2}, which returns a similarity score between 0 and 100 for two strings, where a score of 100 indicates that the shorter string is a sub-string of the larger string. A threshold value for the returned score of the comparison is chosen empirically, indicating when an entry is marked as a duplicate.

If two entries contained identical street number and province information, then their street names and facility names were compared. When these were nearly identical (defined as having the sum of the similarity scores for the facility names and street names to be at least 195 out of a possible 200), then the entries were marked as duplicates. Recognized duplicates were deleted without manual intervention. The chosen threshold was selected close to the maximum score, which minimized any removal of false positives. When duplicates were found, whichever record contained more non-empty fields was retained. In total, 2,435 duplicates were removed.

Identification of Invalid Entries

A pair of filters was used to process the data after the address parsing stage. This captured entries with invalid postal code or province code information and wrote them to a file separate from the database for further processing. Most of these entries were manually corrected and added back into the database. The choice of these two filters is based on their capabilities in detecting potential errors in postal codes and province codes.

Other Data Cleaning Steps

• Data entry formatting (removal of excess whitespace and punctuation), removal of postal code, province/territory names.
• During processing, separation of entries with incorrect postal code or 2-letter province/territory code format from the cleaned data and their manual editing.

Selection of Record to Retain in Case of Duplicates

In some instances, a facility was present in more than one source. In such cases, the record with the most information available was retained. Where information between sources did not match, validation tools were used to decide which to retain.

Classification Used and Assignment of Cultural and Art Facility Type

The original data sources use a variety of standards, classifications and nomenclature to describe the type of cultural and art facility. Unfortunately, there is no classification for cultural and art facilities in Canada that is used universally. The following classification of cultural and art facilities is used for Version 1.0 of the ODCAF:

• Arts or cultural centre: Establishments primarily engaged in promoting culture and arts
• Artist: Individual artists engaged in creating artistic works
• Festival site: Sites on which arts or cultural festivals are held
• Gallery: Establishments primarily engaged in the display of artistic works
• Heritage or historic site: Sites of cultural, artistic, or historic significance
• Library or archive: Establishments primarily engaged in the display, curation, and sharing of primarily written material such as manuscripts, periodicals, and other items such as maps or images
• Miscellaneous: Establishments associated in some way with promoting or providing culture or arts that do not fall into any of the above categories
• Museum: Establishments primarily engaged in the display, curation, and sharing of collections of artifacts, fine arts, and other objects of artistic, cultural, or historical importance
• Theatre/performance and concert hall: Establishments primarily engaged in the public performance of artistic or cultural works

The classification is intended to have broad categories that are helpful in distinguishing major types of facilities and yet enable accuracy in mapping source-specific facility types. Facility types are determined from source-specific facility types and source coverage metadata information. Assignments are made using keywords and validated afterwards, with changes made manually whenever needed. When classifying facilities based on source metadata information, this was done analytically on a case by case basis.

Geocoding and Determination of Census Subdivision

In general, the data included in the ODCAF are what is available from the original sources without imputation. The exception to this is the geocoding and the imputation of CSD names and categories, discussed below.

Census subdivision (CSD)^{Footnote 3} names were derived from two different attributes in the data.

The first attribute comprises the geographic coordinates, namely latitude and longitude. These are placed into the corresponding CSDs by linking the coordinate points to the CSD polygons through a spatial join operation using the Python package GeoPandas.^{Footnote 4}

The second attribute is the city name, where literal string matching was done with each cultural and art facility municipality name and a list of CSD names. The city names with at least ten entries that did not receive a CSD name through this process were manually assigned a CSD name by using Place Names in GeoSuite.

Geocoding was carried out for some sources that provide address data but no geo-coordinates. Latitude and longitude were determined and validated using tools on the internet. A subset of the source-provided geo-coordinates were also validated using the internet. Some coordinates have also been removed from the original sources when it was determined they were derived from postal codes or other aggregate geographic areas as opposed to street address.

While efforts have been made to ensure the accuracy of geo-coordinates, no guarantees are implied, and errors and inaccuracies are possible.

Inclusion in the ODCAF of Facility Type Provided in Source Datasets

The facility types as provided in the data sources (e.g., exhibition or cultural centre, community library, centre d'art, etc.) are also included in the ODCAF without any modification, reassignment, or mapping to a uniform classification.

6. Database Coverage

The ODCAF current version (Version 1.0) database as provided contains approximately 8,000 cultural and art facilities.

As the total number of all cultural and art facilities in the country is not known with a reasonable degree of certainty, the coverage obtained with the sources used was not quantitatively assessed. However, many of the sources purport to list all facilities of a certain type within a jurisdiction. Thus, within these facility type categories and jurisdictions, coverage would be expected to be fairly complete. However, if facilities of a certain category were omitted in a source, then these might be missing from the database, unless they were obtained from a different source.

7. Data Quality

All cultural and art facility data in the ODCAF were collected from government data sources, either from open data portals or publicly-available webpages. In general, other than the processing required to harmonize the different sources into one database, the underlying datasets were taken "as is." The accuracy and completeness of the information is in general a function of the source datasets used.

Classifying facilities

Assignment of facility type was largely based on facility types provided by source datasets. In instances where facility type was either unclear or not defined by the source, facility type was classified based on further research or using meta-information, such as name of dataset.

Removing duplicates

Some source datasets do overlap; datasets which cover only a particular type of arts or cultural facility for an entire province, for example, may overlap with data provided only for specific towns. Although deduplication techniques are used, not all duplicates might have been removed. Modifying the deduplication methods to seek out the remaining duplicates would generate numerous false positives, which would require additional manual intervention. Further details are available in the sub-section Removal of duplicates above.

Correcting invalid entries

A few entries with erroneous province/territory names and postal codes were detected and manually corrected. Further details on the identification of erroneous entries are also reported in the sub-section Identification of invalid entries above.

Address parsing

Natural language processing methods were used for parsing and separation of address strings into address variables, such as street number and postal code (which is removed from the final released database). The methods are reputable in the field for performance and accuracy, but as with all statistical learning methods, they have limitations as well. Poor or unconventional formatting of addresses may result in incorrect parsing. At this stage, no further integration with other address sources was attempted; hence, although address records are generally expected to be correct, residual errors may be present in the current version of the database.

8. Data Dictionary

This data dictionary below describes the variables of the ODCAF.

Arts and cultural facilities varables

Variable – Index

Variable – Facility Name

Variable – Source Facility Type

Variable – ODCAF Facility Type

Location Variables

Variable – Unit Number

Variable – Street Number

Variable – Street Name

Variable – City

Variable – Province/Territory

Variable – Province Unique Identifier

Variable – CSD Name

Variable – CSD Unique Identifier

Variable – Longitude

Variable – Latitude

Variable – Data Provider

9. Contact Us

The LODE open databases are modelled on ongoing improvement. To provide information on additions, updates, corrections or omissions, or for more information, please contact us at statcan.lode-ecdo.statcan@statcan.gc.ca. Please include the title of the open database in the subject line of the email.

Catalogue number: 21260001
Issue number: 2020001

The Open Database of Cultural and Art Facilities (ODCAF) is a collection of open data containing the names, types, and locations of cultural and art facilities across Canada. It is released under the Open Government License - Canada.

The ODCAF compiles open and publicly available data on cultural and art facilities across Canada. Data sources include provincial/territorial governments, municipal governments, and professional associations. This database aims to provide enhanced access to a harmonized listing of cultural and art facilities across Canada by making it available as open data. This database is a component of the Linkable Open Data Environment (LODE).

Data sources and methodology

The inputs for the ODCAF are datasets whose sources include provincial, territorial and municipal governments, and professional associations. These datasets were available either under one of the various types of open data licences, e.g., in an open government portal, or as publicly available data. Details of the sources used are available in a 'Data Sources' table located within the downloadable zipped ODCAF folder.

The data sources used do not deploy a uniform classification system. The ODCAF harmonizes facility type by assigning one of nine types to each facility. This was done based on the facility type provided in the source data as well as using other research carried out for that purpose.

The facility types used in the ODCAF are: art or cultural centre, artist, festival site, gallery, heritage or historic site, library or archive, museum, theatre/performance and concert hall, and miscellaneous.

The ODCAF does not assert having exhaustive coverage and may not contain all facilities in scope for the current version. While efforts have been made to minimize these, facility type classification and geolocation errors are also possible. While all ODCAF data are released on the same date, the dates as of which data are current depends on the update dates of the sources used.

A subset of geo-coordinates available in the source data were validated using the internet and updated as needed. When latitude and longitude were not available, geocoding was performed for some sources using address data in the source street address.

Deduplication was done to remove duplicates for cases where sources overlapped in coverage.

This first version of the database (Version 1.0) contains approximately 8,000 records. Data were collected by accessing sources between January 2020 and July 2020.

The variables included in the ODCAF are as follows:

• Index
• Facility Name
• Source Facility Type
• ODCAF Facility Type
• Provider
• Unit
• Street Number
• Street Name
• Postal Code
• City
• Province or Territory
• Source-Format Street Address
• Census Subdivision Name
• Census Subdivision Unique Identifier
• Province or Territory Unique Identifier
• Latitude
• Longitude

For more information on how the addresses and variables were compiled, see the metadata that accompanies the ODCAF.

Downloading the ODCAF

For ease of download, the ODCAF is provided as a compressed comma-separated values (CSV) file.

• ODCAF Version 1.0

Visualizing the ODCAF

The ODCAF content is available for visualization on a map using the Linkable Open Data Environment Viewer.