HomeMy WebLinkAboutData Governance StandardsData Governance Standards – August 2020 1
City of Chula Vista
Data Governance Standards
Updated August 2020
Data Governance Standards – August 2020 2
Contents
Introduction ........................................................................................................................................................................ 3
Part 1: Data Governance Roles & Responsibilities ............................................................................................................. 4
Part 2: Data Inventory Annual Process ............................................................................................................................... 6
Step 1. Identify/verify city datasets ................................................................................................................................ 7
Step 2. Identify/verify data-sharing relationships and complete data “wish list” .......................................................... 7
Step 3: Review and prioritize datasets ........................................................................................................................... 7
Part 3: Publishing Open Data .............................................................................................................................................. 9
Step 1. Assess sensitivity .............................................................................................................................................. 10
Step 2: Mitigate risks .................................................................................................................................................... 11
Step 3: Upload .............................................................................................................................................................. 12
Step 4: Assess quality and publish ................................................................................................................................ 12
Part 4: Updating City of Chula Vista Data ......................................................................................................................... 14
Appendix A: Data Inventory Worksheet
Appendix B: Data Sharing Worksheet
Appendix C: Recognizing Sensitive and Protected Data
Data Governance Standards – August 2020 3
Introduction
The City of Chula Vista recognizes the data it generates as a strategic asset. City data is not merely a byproduct of the
City operations, but a valuable tool that can be used to generate insight, align priorities, and make decisions.
Through the Smart City Strategic Action Plan and the Open Data Policy, the City Council has prioritized a shift toward
making City-owned data easier to access and analyze both internally — among different teams and departments — as
well as externally with other government agencies, nonprofit organizations, academic researchers, and residents at
large.
Providing access to city data through an open data program creates value by providing the public with transparency
and a better understanding of City operations. Open data helps break down silos within the City organization and
between the City and other agencies, potentially increasing the accuracy and effectiveness of policies and programs.
Open data can also improve and strengthen the dialogue between the City and the public, building trust, promoting
accountability, and establishing a shared sense of priorities.
The City also recognizes that the opportunities presented by open data come with challenges and risks. Protecting
individual privacy is of paramount concern when data is released to the public. Ensuring that open data is accurate,
current, and in accessible formats is essential to its value. Presenting the data in its proper context is critical to
facilitating understanding and avoiding misinterpretations and improper use.
The City has established a Data Governance Committee for the purpose of navigating these challenges and risks and
delivering value from open data. The Data Governance Committee, comprising representatives from various City
departments, is responsible for setting priorities, protecting privacy, ensuring accuracy, resolving conflicts, and
directing the design and implementation of data portals and visualization tools.
It is important to note that while the Data Governance Committee performs high-level leadership and oversight, the
success of the open data program requires the involvement of many City employees and external stakeholders. In
each City department, staff who work directly with data are key to the program’s ability to function. Described as
data coordinators, these staff are the experts on specific datasets and serve as the stewards of that data in the
context of the open data program.
This Data Governance Standards document serves as the primary manual and policy guide for the open data program.
Developed by the Data Governance Committee with insight from successful open data programs in other U.S. cities
and agencies, these standards form the basis of the program. This document will be expanded and updated
periodically based on new developments in the program.
Data Governance Standards – August 2020 4
Part 1: Data Governance Roles & Responsibilities
The Data Governance Committee includes the Program Lead and a representative selection of City staff who support
the program with decisions and policies that require specialized knowledge of City operations, legal matters, or
technical systems. The Data Governance Committee makes high-level decisions about data-sharing agreements, open
data, data publication priority, and other related issues.
The Program Lead is responsible for the implementation of the vision for the open data program. The Program Lead
guides the work of the Data Governance Committee to set priorities, define related policies, and coordinate with the
open data program’s technical systems and organizational processes.
Data Coordinators are the subject matter experts in their respective service areas and will help identify potential
datasets, upload datasets to the open data portal, and periodically update data. Data Coordinators serve as a liaison
between the Data Governance Committee and their respective departments or service areas. Data Coordinators may
also participate directly on the Data Governance Committee. Every department, including Police and Fire, has at least
one designated data coordinator.
Figure 1 and Table 1 outline in more detail the specific roles and responsibilities of various City staff in the context of
data governance. Note that some individuals might serve more than one function.
Figure 1: Data Governance Relationships
City
Management
Data Governance
Committee
Program Lead IT / GIS Data
Coordinators
Communications
Manager
City Attorney
Data Governance Standards – August 2020 5
Table 1: City Staff Roles and Responsibilities
Role Description Chula Vista Staff
Executive Support • Executive level support and encouragement of the open data
process. This is critical to gaining buy-in citywide.
• Publicly announce the new open data program or policy, as well as
provide encouragement and support to the City’s Data
Governance Committee.
City Management
Strategic
Alignment • Responsible for overall program leadership; ensures alignment
with overall city and administration goals.
• Create accountability for department heads and delegated
departmental staff; help set and approve priorities for data
releases; reports to the City Manager or Assistant City Manager(s)
on program activities.
• Assign data coordinators within each service area.
Open Data
Program Lead
Communications
Strategy • Ensure public datasets are presented with messaging consistent
with city goals and programs.
• Assist city leadership with potential responses to public input on
published dataset.
• Oversee public engagement activities and events.
Communications
Manager
Technical
Operations • Oversee the technical aspects of open data, such as maintaining
the public data catalog, extracting data from internal IT systems,
and ensuring public data is kept up to date.
• Build and configure the tools needed to routinely copy data from
city technology systems to the open data website.
IT Director
GIS Manager
Data Coordinators • Work with the program lead and others to complete periodic data
inventories and ensure the publicly released data is accurate, well
documented, and up to date.
• Review source data systems; recommends specific data elements
for release approval; helps develop public documentation; handles
manual data publishing if needed.
Department staff
as assigned
Legal Strategy • Assist in developing policy and guidance materials for releasing
datasets.
• Develop terms of service and licensing for data access and usage.
• Resolve legal considerations for complex datasets.
City Attorney’s
Office
Data Governance Standards – August 2020 6
Part 2: Data Inventory Annual Process
The City maintains a comprehensive inventory of the various sets of data in its possession. Maintaining this catalog of
datasets helps City staff and the public understand what data might be available to them.
A data inventory is a fully described record of the various sets of data maintained by the City. The inventory includes
basic information about each dataset, such as its name, a description of its contents, how frequently it is updated,
who owns and maintains the dataset, whether it is subject to privacy laws, the source of the data, and other relevant
details. These details about a dataset are known as metadata.
It is important to keep in mind that the data inventory does not consist of any raw data. The data inventory provides
metadata only. The actual contents of the datasets are maintained separately from the data inventory.
Each year, the City updates the data inventory using a prescribed process to identify datasets, identify data-sharing
opportunities, and prioritize datasets for publication in a public open data portal.
The process of updating the data inventory is led by the open data program lead and the Data Governance
Committee, with support from data coordinators in each of the city’s departments and program areas.
Figure 2: Data Inventory Process
• Data coordinators
complete inventory
worksheet
• Data coordinators
identify which of their
datasets are shared
with other
departments
• Data coordinators
identify which datasets
they want or need from
other departments
• Data Governance
Committee checks
worksheets for
accuracy and
completeness
• Data Governance
Committee prioritizes
datasets based on
demand and difficulty
of implementation
For a data inventory worksheet template, see Appendix A. For a data sharing worksheet template, see Appendix B.
Identify / verify
datasets
Identify/verify
data-sharing
relationships
Review and
prioritize
datasets
Data Governance Standards – August 2020 7
Step 1. Identify/verify city datasets
Each fiscal year, data coordinators around the City work with their departments to complete an inventory worksheet
that lists all datasets created or managed by that data coordinator’s respective service area. The existing data
inventory, as well as worksheets from the previous year, may be used as a starting point for the following year’s
inventory.
In addition to identifying datasets and providing details for each dataset, the worksheet asks data coordinators to
suggest a draft priority level for each dataset. This suggested priority level reflects the data coordinator’s opinion on
how valuable or impactful the data might be to another user.
For a sample worksheet, see Appendix A.
Note that all datasets are included in the inventory, including datasets that may contain sensitive or restricted data,
or datasets that may not seem valuable to the open data program.
Step 2. Identify/verify data-sharing relationships and complete data “wish list”
While completing the data inventory, data coordinators also identify which datasets are regularly being shared
among departments, and identify which departments regularly receive a given dataset. The purpose of tracking these
data sharing relationships is to identify opportunities to create efficiencies and assist in the prioritization of data
publication.
In addition to noting which datasets are regularly being shared among departments, the worksheet asks data
coordinators to make a “wish list” specifying datasets that their department or team wishes to procure from another
department.
For a sample worksheet, see Appendix B.
Step 3: Review and prioritize datasets
Following the submission of data inventory worksheets, the program lead coordinates with the Data Governance
Committee to consolidate the various worksheets into a master inventory.
The program lead, Data Governance Committee, and IT team then perform a review and gap analysis to identify any
datasets missed by data coordinators. Chula Vista employs a series of relational databases to house complex datasets,
and data within these important databases could be missed by some coordinators. The program lead works with the
IT team to conduct a gap analysis between the results of the data inventory and the contents of the City’s various
relational databases. Data coordinators then assist with filling in any missing details.
The data coordinators will have suggested draft priority levels for datasets in their respective worksheets. The
program lead, in consultation with the Data Governance Committee, will review these draft priority levels and make
adjustments as necessary.
The priority level will be based on two dimensions, the demand for or value of the dataset, and the ease of
publication. Priority levels will be assigned as high, medium, or low in each area. Data highest in demand and lowest
in difficulty of implementation will be considered Priority 1 (P1). Data lowest in demand and highest in difficulty of
implementation will be considered Priority 4 (P4). Table 2 shows the priority levels resulting from the relationship
between demand and difficulty of implementation.
Departments will review the draft inventories and priority levels and have an opportunity to submit corrections.
Data Governance Standards – August 2020 8
Table 2: Dataset Prioritization Matrix
Demand/Value:
• High – The dataset is tied to a performance management goal, a City Council or City Manager’s initiative, or
would provide immediate benefit to multiple departments, the organization, or the community for making
data-informed decisions.
• Medium – At least one of the following is true concerning the dataset. The dataset has been previously
requested or has the potential to be requested by the public or business community. The dataset has been
included on a department data wish list as defined in the previous section.
• Low – The dataset historically has had little or no public demand. Little organizational, departmental,
business, or public benefit has been identified.
Difficulty of implementation:
• Low – Most of the following are true concerning the dataset. Very minor (if any) sensitivity concerns are
found in the dataset and can be quickly and easily mitigated by the data coordinator and program lead.
Metadata is already accessible, or easy to produce. The data is already in a format that can be uploaded to
the portal or is easily configured and converted. An existing automation process used to maintain data on the
portal can be implemented for the dataset.
• Medium – The dataset configuration will require a concerted effort but can be generated in a format that the
portal accepts without deviating from the established business process. Sensitivity concerns are minor and
successfully addressed by the data coordinator and program lead. Metadata can be generated for the dataset
within the current business process for creating open datasets.
• High – One or more of the following is true concerning the dataset. The dataset contains significant
sensitivity concerns that will require an extended sensitivity review process, a legal department review, or a
significant revision of the dataset to resolve. The dataset requires significant configuration to generate the
file in a format the portal accepts. Setting up the automation process will require a change in business
process and will require significant effort.
HighP2 P3 P4
MediumP2 P2 P3
LowP1 P2 P2
High Medium LowDifficulty of implementationDemand / Value
Data Governance Standards – August 2020 9
Part 3: Publishing Open Data
After the data inventory and prioritization process is complete, datasets are uploaded to the open data portal using
the following process to effectively ensure strict adherence to data privacy protections and legal requirements.
Figure 3: Data Publishing Process
Data Governance Standards – August 2020 10
Step 1. Assess sensitivity
Transparency is a core goal of Chula Vista’s open data program, but transparency must be balanced with the
imperative of protecting sensitive and private information.
The City’s Open Data Policy (Policy Number 112-03) excludes the following protected data from publication:
• Any data which if published would violate state or federal laws governing privacy or privileged information
• Any data which if published would violate contractual or proprietary rights
• Any data which if published would jeopardize the security of individuals or property
Refer to Appendix C for further discussion and examples of potentially protected data.
For each new dataset that has been identified for publication, the data coordinator and program lead will perform a
sensitivity assessment to address concerns about protected data. The data coordinator begins by documenting any
sensitivity concerns and sharing these concerns with the program lead. The program lead reviews these concerns and
classifies the dataset as having “no concerns,” “minor concerns,” “significant concerns,” or “overwhelming concerns.”
If there are no concerns or only minor concerns, the data coordinator is responsible for applying appropriate
mitigation (see Step 2). Once the data coordinator has applied mitigation measures, the program lead will verify that
all concerns have been addressed and will present the concerns and mitigations to the Data Governance Committee
for approval. If the data coordinator does not complete the mitigation within 30 days, the dataset will automatically
be forwarded to the Data Governance Committee for review.
In cases of significant or overwhelming concerns, the program lead will present the dataset to the Data Governance
Committee, which will determine appropriate mitigation measures to be applied by the program lead. The committee
will confer with the City Attorney’s office if necessary to ensure any proposed mitigation adequately resolves
concerns. Table 3 provides more detail about this process.
Data Governance Standards – August 2020 11
Table 3: Sensitivity assessments
Level of concern What it means How it is handled
No concerns The dataset does not contain any
information that a reasonable person
would consider protected.
Skip to Step 3, Upload.
Minor concerns The dataset contains some protected
data, but these data can be easily
mitigated (for example, a column that
can be easily removed from the dataset).
These concerns are communicated to the
data coordinator or program lead for
mitigation (see Step 3). The dataset is then
re-reviewed by the program lead and the
Data Governance Committee.
Significant concerns The dataset contains protected data that
cannot be easily resolved, but the issue
does not immediately disqualify the
dataset for publication (for example, data
is anonymized but could be easily re-
identified when combined with another
published dataset).
The dataset and documented concerns
are forwarded to the Data Governance
Committee and, if appropriate, the City
Attorney’s office, for review and, if
possible, mitigation.
Overwhelming concerns The dataset contains significant
protected data that obviously cannot be
mitigated (for example, the dataset is
clearly protected by state law).
The dataset is presumptively disqualified
from publication but forwarded to the
Data Governance Committee and City
Attorney to determine if any portion of
the dataset can be published without
compromising protected data or violating
the Open Data Policy.
Step 2: Mitigate risks
If there are any concerns about protected data identified during Step 1, then the data coordinator, program lead
and/or Data Governance Committee will identify and apply appropriate risk mitigation techniques, which may
include:
• Redaction (e.g., deleting a column of names).
• Anonymization (e.g., reducing the precision of addresses).
• Aggregation (e.g., averaging a dataset by age group).
• Constraining access (e.g., limiting access to city staff).
• Forgoing publication and discarding the dataset.
Redaction is the process of removing sensitive information from a dataset. In the context of open data, redaction
generally means deleting entire columns or rows from the dataset. It may also mean deleting specific cells if those
cells contain sensitive data.
Anonymization is the process of removing sensitive information and replacing it with a value that cannot be tied to
an individual person’s identity. For example, a spreadsheet of all registrants for City recreation programs could be
anonymized by removing any identifying details and replacing the name with a random number. Anonymization could
also mean reducing the specificity of identifying information. For example, a spreadsheet of all police calls for service
could be anonymized by removing all street address numbers and replacing them with block numbers only.
Data Governance Standards – August 2020 12
Aggregation is the process of combining multiple entries within a dataset into one common category. For example, a
dataset of all citations issued for residential code violations could be aggregated so that the totals are only presented
for each Council District, or each ZIP Code, rather than each individual home address.
Constraining access is a less common mitigation measure that may be appropriate for data sets that have high value
but that also present significant or overwhelming sensitivity concerns. Constraining access is the process of making a
dataset available only to selected users. For example, access to the Human Resources department’s timesheet data is
limited to the Finance Department and select others within the City, but this data is never made available to the
public or any external users.
Forgoing publication is the last resort if other mitigations are not available. Some datasets simply contain too much
protected data that cannot be mitigated using one of the methods described above. In these cases, the Data
Governance Committee may decide that the dataset is not a candidate for publication.
While the above-listed mitigation methods are the most common ways of mitigating, there may be other mitigation
measures that are appropriate in specific circumstances.
Step 3: Upload
After completing sensitivity review and mitigation, datasets will be uploaded to Chula Vista’s open data portal.
Upload can proceed in one of two ways:
• Option 1: Data coordinators upload data directly to the portal. This method is ideal for large departments
that have many datasets to contribute to Chula Vista’s open data portal. The open data program lead is
responsible for training these data coordinators on using the City’s open data portal.
o Data coordinators will coordinate the dataset upload with the open data program lead.
o In addition to uploading datasets, data coordinators will input metadata (e.g., readme text, column
descriptions) to contextualize datasets and dataset attributes. Metadata must include information
on update frequency.
o After the data coordinator has uploaded a new dataset and entered metadata, he or she will share
the private dataset with the program lead for review.
• Option 2: Data coordinators send data to the open data program lead, or the program lead’s designee, for
upload to the portal. This method is intended for smaller departments and groups that may have only a small
number of datasets to contribute to Chula Vista’s open data portal.
o Along with new datasets, owners will send metadata (e.g., a readme text file, column descriptions) to
contextualize datasets and dataset attributes. Metadata must include information on update
frequency.
o The program lead or their designee will upload new datasets, input metadata, and then share the
private dataset with data coordinators for review.
Step 4: Assess quality and publish
After a dataset and associated metadata have been uploaded to the portal, but before they are published, a final
quality assessment (QA) review takes place to find and correct any errors in the data itself or in the metadata. The QA
is conducted by either the program lead or the data coordinator (whoever did not upload the data, as the intent is to
subject the dataset to a fresh set of eyes).
Datasets that pass both the sensitivity review and QA are made public on the portal.
Each dataset is unique and requires a different process to review and assess quality. However, certain commonalities
apply across all datasets. Chula Vista’s open data program employs the following quality standards during the dataset
QA process:
Data Governance Standards – August 2020 13
• The dataset is the most complete, accurate, and current version appropriate for public release.
• The data have been spot-checked for common errors such as missing and misplaced values.
• Any missing data points are left as null, but the meaning of null is defined in the dataset’s metadata.
• Columns are formatted appropriately.
• Metadata is complete, concise, and free of jargon.
• Metadata explain the process used to create the data and summarize any changes.
• Metadata clearly explain any limitations or omissions for each dataset.
• Metadata clearly identify an update frequency and plan.
Before publishing a new dataset (or updating an existing dataset), the reviewer (program lead or data coordinator)
must verify whether any existing dashboard or other tool is dependent on that dataset. The reviewer must coordinate
with whomever is responsible for the associated dashboard or other tool to ensure that updating or replacing the
dataset will not interfere with the intended operation of the dashboard or tool.
It is important to note that the open data program is not a data editing program. No dataset is perfect. Data
coordinators will not be asked to substantially change or improve datasets pulled from city systems before uploading
them to the portal. Data coordinators are not asked to impute missing values, though they may do so if they wish.
The primary goal of the program is not to improve city data, but rather to share existing data with the public and
among city staff. The program also provides an opportunity to gain feedback about the quality of data from other
departments and the public. Any improvement to internal processes for data creation and curation that results from
the open data program or associated feedback will be considered welcome but auxiliary.
Data Governance Standards – August 2020 14
Part 4: Updating City of Chula Vista Data
A successful open data program requires current, up-to-date data. When submitting a dataset to be published to the
portal, data coordinators must ensure that update frequency is included in the metadata. Coordinators are
responsible (with assistance, if needed, from the program lead) for all manual updates to datasets. Whenever
possible, coordinators will update an existing dataset instead of creating a new dataset.
Nevertheless, data coordinators may have limited time to monitor the freshness of datasets posted on the portal. The
program lead will employ two tactics to help ensure the data portal remains up to date:
• Active monitoring of dataset freshness. The program lead will periodically review each dataset on the portal
and will alert data coordinators when a dataset is identified as out of date and needs to be refreshed.
• Automation. For datasets that need to be updated at least once per month or more, the program lead will
work with data coordinators and the IT database administrator to implement automatic updates, where
appropriate.
Appendix A
Appendix A: Data Inventory Worksheet
Dataset Title Department Division / Team Data Owner Description Format System / Source Sensitivity Sensitivity Comments Update Frequency Data Quality Data Value Accessibility Draft priority
Human-readable name
of the asset. Should be
in plain English and
include sufficient detail
to facilitate search and
discovery. Avoid
acronyms.
Department where the
data is maintained. If
more than one
department, list the
primary responsibility.
The name of the
division or team within
the department where
the data is maintained.
Name of the staff
member who is
primarily responsible
for maintaining the
dataset.
What the dataset
describes. Provide a
longer description of
the data that can be
readily understood by
non-technical users.
What file format is the
data in? Examples: CSV,
ArcGIS Feature Class
What is the original
system/application/file
that houses it? Please
provide additional
comments as
necessary.
Does the dataset
contain sensitive data,
such as personal phone
numbers, Social
Security Numbers, etc.?
What could be done to
make the data
shareable?
How often is the
dataset updated?
Examples: Hourly, Daily,
Weekly, Quarterly
How accurate and
consistent is the data?
How valuable is the
data to the city
internally?
How difficult is it to
access and publish this
data?
Page 1 of 1
Appendix B
Appendix B: Data Sharing Worksheet
Dataset Title Owner Department Owner Division/Team Requesting Department Requesting Division/Team Description of Use Update frequency requested Current contact Notes
Human-readable name
of the asset. Should be
in plain English and
include sufficient detail
to facilitate search and
discovery. Avoid
acronyms.
Department where the
data is maintained. If
more than one
department, list the
primary responsibility.
The name of the
division or team within
the department where
the data is maintained.
Department where the
data is used or will be
used.
The name of the division or
team within the department
where the data is used or will
be used.
Describe the purpose
for sharing the data /
how it is used by the
requesting department
How often does the dataset
need to be updated to be useful
to the requesting department?
Page 1 of 1
Data Governance Standards – Appendix C – August 2020 1
Appendix C: Recognizing Sensitive and Protected Data
The City’s Open Data Policy (Policy Number 112-03) excludes the following protected data from
publication:
• Any data which if published would violate state or federal laws governing privacy or privileged
information
• Any data which if published would violate contractual or proprietary rights
• Any data which if published would jeopardize the security of individuals or property
To assist staff in recognizing data that may be sensitive or protected from publication, this appendix
provides examples and discussion from several sources.
The Chula Vista City Clerk’s Office, in evaluating whether data is exempt from disclosure under the
California Public Records Act, considers the following parameters as part of its decision on what to
release publicly:
“Protected Data” is any Data or Dataset:
• that is otherwise exempt from disclosure pursuant to California laws including but not limited to
the California Public Records Act.
• that contains a significant amount of Data that, if disclosed, would impose an undue financial or
administrative burden on the City.
• that reflects the internal deliberative or administrative process(es) of the City, including, but not
limited to, Data relating to negotiating positions, future procurements or pending or reasonably
anticipated legal or administrative proceedings.
• that is subject to privacy laws or subject to copyright, patent, trademark or trade secret
protection, or to a confidentiality agreement, or that are otherwise protected by law or
contract.
• that includes or constitutes proprietary applications, computer code, software, operating
systems or similar materials.
• that includes or constitutes employment records, internal employee-related directories or lists,
protected health information.
• which, if disclosed by the City, might in the City’s discretion, raise privacy, confidentiality or
security concerns or jeopardize or have the potential to jeopardize public health, safety or
welfare.
When making data available, staff should consider the possibility that multiple data sets could be
combined, resulting in the inadvertent availability of protected data and take steps to ensure the
protection of such data.
The following table shows exemptions to consider when producing data/datasets for common data and
many of the most frequently raised exemptions from disclosure that may, or in some cases, must be
asserted by local agencies:
Data Governance Standards – Appendix C – August 2020 2
Type of Data Example Data Legal Citation
Incomplete Datasets Incomplete Data; Preliminary Information;
Drafts; Notes
Government Code §6254(a)
Personal Contact
Information
Address; Phone Number; Email Address Government Code 6255
Licensee Financial
Information
Personal financial data required by
applicants
Government Code §6254(n)
Law Enforcement Arrest records or complaint reports;
investigatory or security files;
intelligence information
Government Code §6254(f)
Library Circulation Data kept for the purpose of identifying
the borrower of items available in
libraries.
Government Code §6254(j)
Pending Litigation/Claims Pertaining to pending litigation or
unsettled claims
Government Code §6254(b)
Personnel, Medical or similar
files
Employment records; Questions scoring
keys and other examination data
Government Code §6254(c)(g)
Real Estate Appraisals; Engineering Evaluations Government Code §6254(h)
Taxpayer Information Data required from taxpayer in connection
with the collection
of local taxes
Government Code §6254(i)
Trade Secrets/Proprietary
Information
Records and information related to the
award of a public
contract.
Government Code §6254.15
It is important to realize that while entire datasets may clearly be of a sensitive nature, it is also possible
for smaller subsets of data within a larger dataset to present problems. For example, a single row or
column within an otherwise benign dataset could contain sensitive data. A single cell, or handful of cells,
within a dataset, may contain unexpected sensitive data, which may have been placed there either
intentionally or by mistake.
It may be helpful to think of sensitive data as falling into one of several categories.
Direct or unique identifiers: Sometimes known as Personally Identifiable Information (PII), these
attributes or features can clearly identify individuals. Examples:
• Name
• Social security number
• Employee ID numbers
• Email address
• Home address
• Phone number
• Driver’s license number
Indirect or quasi-identifiers: Attributes or features within a dataset that do not clearly identify
individuals on their own; however, in combination with other data, these attributes can lead to the
identification of an individual in what is known as the “mosaic effect.” Examples:
Data Governance Standards – Appendix C – August 2020 3
• Birth date
• Zip code
• Gender
• Race / ethnicity
• Income level
• Age or age bracket
• Medical conditions
Addresses: Street addresses are an often-overlooked source of sensitive data. While home addresses
can obviously be unique identifiers, other addresses or place names can act as quasi-identifiers that
reveal sensitive details about individuals when combined with other quasi-identifiers. Additionally,
geographic coordinates can function in the same way as street addresses.
Unstructured fields: Unstructured fields contain open-ended written comments or descriptions (as
opposed to structured fields, which only offer a limited range of pre-determined values). Examples
include:
• 9-1-1 call descriptions
• Online survey responses
• ACT Chula Vista service requests
• Any other freeform text fields that could be filled in unpredictable ways
If you are unsure if data should be considered sensitive or protected, discuss it with the Open Data
Program Lead, who may also share the question with the Open Data Governance Team.
This guide was developed with significant reference to the following sources:
• DataSF Open Data Release Toolkit: Privacy Edition:
https://docs.google.com/document/d/1_K59q9ik5eEw9_-iiPCbQ8WSlGk0FovNz1uLajVHoi0/edit
• “Open Data Privacy” by Ben Green, et al., Berkman Klein Center for Internet & Society at
Harvard University:
https://dash.harvard.edu/bitstream/handle/1/30340010/OpenDataPrivacy.pdf