As we embrace artificial intelligence (AI) advancements, I am asking that everyone please take a minute to review our��data secure policy, which is vital to keep our student, patient, and other University data protected. It is imperative to remember that non-public or sensitive University information should never be uploaded into external AI tools—whether free or paid—unless there is a university agreement with the vendor approved by one of the various AI governance groups. This policy is crucial for maintaining our data security and applies to all members of our community.

I encourage you to take advantage of our secure version of a generative AI chatbot, designed to securely handle medium- and high-risk institutional data. Accessible at��, this advanced AI tool is exclusively for our faculty, staff, and students, providing a reliable and secure platform for managing a variety of inquiries and tasks. I am also excited to inform you that additional features will be introduced soon, further enhancing its capabilities. Please note that while the site is internet accessible, two-factor authentication (Duo) is required when off-network.

To begin using our secure AI chatbot, please visit����and log in with your UR or URMC credential.��Help with logging in is available here.

Should you require any assistance or have questions, the University IT or ISD Help Desk are available to support you at��univithelp@rochester.edu��(585) 275-2000 or��ISDHelpdesk@URMC.rochester.edu��(585) 275-3200.

Thank you for your commitment to upholding the integrity and security of our institutional data. Together, we are advancing toward a more secure and Boundless future.

Best regards,

Nicole Sampson
Provost��and Chief Academic Officer
University Professor of Chemistry

The post Responsible Use of AI Tools and Institutional Data Security appeared first on Office of the Provost.

We recommend that standards be adopted when new processes or platforms are developed or existing ones undergo significant updates.

Questions can be addressed to the DGSO.

Organizational Naming Standard

Purpose: 

• The short name and single token name  standards are to be used  for interoperability and consistency for labeling within university business processes.
• For externally shared print and online publications, refer to the University style guide.

Standards: 

Files for Download

• Organization Official Name (text file)
• Organization Official Name Alternative(text file)
• Organization Short Name (text file)
• Organization Single Token Name (text file)
• Organization All Official Name fields (tab delimited file)
• Organization All Alternative Name fields (tab delimited file)

The post Introducing Data Standards: Organizational Naming appeared first on Office of the Provost.

The principles we uphold define our values. Data responsibility is achieved when we take actions that support the data values we believe in. These values can be found on the Data Governance and Support Office website.

Here are some tricks and tips to help you use data in a responsible way:

• Ensure that you are not using data for personal use.
  • Clarify why or how a data point will be used. What is the justifiable purpose for collecting?
• Ensure you are always documenting your work.
• Ensure you are communicating data efficiently
• Don’t utilize data just because you can. Make sure your data usage is meaningful and consistent.
• Explore the ethical usage of data.
• Do not share data with those not authorized to see it or use it.
• Ensure that if you are utilizing data in your work, that you understand the data you are using, know how to explain it via reports, executive summaries, and presentations, and that you can present it to different audiences at different levels.

Read more about using data meaningfully in the .

The post Using Data Responsibly appeared first on Office of the Provost.

Data consistency can be achieved by standardizing an approach to documenting reports, analytics, and statistics to ensure data can be replicated and explained. That is, when doing something such as analyzing or querying data, the formulas are documented and understandable. If there are discrepancies in data, ensure that individuals are able to explain why things are different.

Here are some tricks and tips to help you use data consistently:

• Ensure data systems being utilized are up-to-date and producing correct data.
• Document data systems being utilized and reason for utilizing these systems.
• Document any filters utilized in the data, including filters for dates/snapshots.
• Document each calculation made to the data or data variables and why.
• Document any joins done to the data and how they were defined.
• Document criteria for including and excluding variables or records in the data.
• Follow style guides and standards, as available, for the output of data.
• Define the data variables in the data. Ensure they are either vetted or will go through a vetting process prior to being published if not previously vetted.
• Document and describe how often data will be refreshed, which variables are or are not refreshed within the data, and how and how often data are vetted.
• Describe who the audience for the data are and who is allowed to have access.
• Define and explain any differences in the data if they are being counted in a new or distinctive way.

Read more about using data consistently in the report.

The post Consistent Data Use appeared first on Office of the Provost.

Data become meaningful when we can clearly articulate the problem that it will help solve. Data become meaningful to the University when they are used to answer questions that will satisfy strategic priorities and further the mission of the organization as a whole; today and into the future.

Here are some tricks and tips to help you use data in a meaningful way:

• When asked for data, make sure to ask probing questions to fully understand the request and how the data will be utilized.
• Document the questions that can be answered with data provided (and which cannot) and how these questions aligns with the priorities and mission of the University.
• Understand and document the purpose for reports, analyses, dashboards, etc.
• Ensure the data that are needed to answer the questions satisfactorily are available and collected for analyses.
• Validate data quality (typically in Consistency Phase) to understand any gaps or errors in data. Document these and, if needed, make updates to ensure data quality.
• Do background research and reading to make sure you are conducting correct analyses on data. Make sure you are utilizing the right statistical models.
• Beware of bias risks in the form of both confirmation bias as well as other hidden bias. Address and document any biases in the data.

Read more about using data meaningfully in the .

The post Meaningful Data Use appeared first on Office of the Provost.

But what does it mean to use data consistently? Let’s start with some definitions.

When data are used consistently, the data are used in a standardized and documented way over time, and across all units of the University, especially to be explainable, fair, accurate, and replicable.

When data are used meaningfully, the data have an articulated important, and useful purpose for supporting University processes, which in turn, drive informed decision making.

When an individual uses data responsibly, it means they recognize they have an ongoing obligation to care for the data as part of their roles and responsibilities. The individual makes ethical, conscientious decisions to act upon data in a way that will seek to improve the consistency and meaningfulness of the data in an effort to improve the mission of the University.��

A full report from the data usage team can be found here:

The post Data Use Guidelines appeared first on Office of the Provost.

• Data Security Classification Policy Implementation project has 5 workstreams that kicked off in December and will be meeting through June on the following activities:
  • Update webpages and documents to reflect new data classification labels.
  • Create communication and change management materials to aid users in the adoption of new data security classifications.
  • Update existing processes and policies to reflect new data security classifications.
  • Create New Processes to support the assignment and review of classifications.
  • Create or update training and educational materials around data security that reflect the new classifications and risk based measures.
• Data privacy policy investigation to benchmark and do foundation work to understand how our peers address data privacy issues. Due to kick off in March.
• Data retention guidelines to benchmark and do a gap analysis to understand best practices around data retention, record retention, and archiving. Due to kick off in February.
• Identify and document critical data elements in the student domain to set priorities for data quality efforts for student data. Due to kick off in April.
• Provide process enhancements and make other changes to the data permission request process to provide additional functionality. This work will be on going as more domains are brought into the data request process.
• Formalize the process and infrastructure to support ad hoc data management starting with data for business unit identifiers for nursing staff at the hospital. This work is underway and is likely to be completed by April.
• Improve training and education for data users and data stewards across the University. This includes restarting the data literacy efforts suspended in Spring 2020 as well as more detailed training for targeted audiences.

If you would like more information about any of this work or would like to make suggestions or participate in any DGSO project, please email the Data Governance and Support Office.

Progress updates will be made regularly as the work progresses.

The post Proposed Data Governance Activities for 2021 appeared first on Office of the Provost.

Our vision is that institutional data are securely accessible, understood, and trusted and are used in a consistent, responsible, and meaningful manner in order to increase interoperability, ensure reliability, and inform decision making in our community.

Goals to help achieve that vision:

1. Manage like a master: Enable and support the collection and management of master data elements for broad community use.   Addresses interoperability, reliability, consistency, secure accessibility.
2. Improve intelligence and involvement: build data literacy for staff across the university including data users, business process owners, and technology providers. Addresses understood, meaningful, reliable.
3. Connect and converse:  support the change that is required to build community where data activities are appreciated and data are appropriately shared. Addresses trusted, responsible, and reliable, and informing.

Activities that help achieve these goals:

1. Create master data models as a foundation for broad-use master data collections.
2. Build data collections that the master data models describe and ensure people and processes are in place to keep collections current.
3. Enable a management and dissemination platform for data stewards to create and update master data for downstream users to consume.
4. Improve data awareness in process design by consulting with project teams before processes and platforms are designed, built, or acquired.
5. Formalize data requirements gathering as part of the technology investment process to ensure appropriate data capture, reporting, and storage capabilities.
6. Increase the data usage and handling skills of staff across the university to reduce anxiety about working with data and improve data processes, products, and output.
7. Provide a foundation for collaboration across various silos and data deserts to help manage expectations and improve trust and data sharing.
8. Create robust change management and communication strategies to bring awareness to the importance of data activities.
9. Connect data users and analysts across the enterprise to build a community and increase trust and reduce unnecessary duplication.

The post UR Data Strategy 2020-21 appeared first on Office of the Provost.

Principle statements:

The University undertakes data policies, processes, and practices that reflect the following values:

• We have an enterprise view: Data are an enterprise asset, owned by the University. Data collection, documentation, and management activities should consider the whole university and aim for solutions that meet the broadest possible needs and handle exceptions separately.
• We support discoverability: Data should be made readily discoverable and accessible along with information on permissions, usage, and restrictions.
• We endorse sharing: In order to maximize usefulness and value of critical assets, data should be made as broadly available as possible subject to legal and regulatory restrictions and following security and privacy policies and standards. The default for sharing information should be “unless there is a good reason not to” rather than “only if you must.”
• We benefit from standards: Standard definitions, naming conventions, and metadata specifications should be used whenever practicable to maximize interoperability, promote consistent use, and provide comparability across data sources.
• We stipulate user rights and responsibilities: In return for data access and support, users agree to undergo necessary training, follow established policies and protocols, and acknowledge that access does not imply unrestricted use. Data users, which includes faculty and staff, are responsible for proper management of data, adherence to rules for usage, and meeting requirements of external authorities, such as regulatory bodies or research granting agencies.
• We promote good data management: Data management responsibilities should be documented as part of relevant job descriptions and appropriate resources be made available to ensure data stewardship. Good data hygiene, including data quality management, is an on-going process and not a one-time project.

The post Data Principles appeared first on Office of the Provost.

Please join us for the following sessions:

• Data Governance: Not the Dark Side of the Force

September 12, 2019

Presenter: San Cannon, Chief Data Officer

Data Governance isn’t just about being told what you can’t do with data.�� It’s a discipline that provides all data management practices with the necessary structure, strategy, and support needed to ensure that data are managed and used as a critical University asset.�� Come learn what the Data Governance and Support Office is doing to help make it easier for University data users to do good things with our data.

(51 minutes, mp4 format, Box login required)

(PDF, Box login required)

• Discovering Admin Data

October 10, 2019

Presenter: John Podvin, Senior University Director for Institutional Research

What data are available to me at Rochester? The University manages a wealth of data. This session will include a panel discussion featuring UR staff responsible for some of our administrative data resources. Panelists will introduce the data that they manage, discuss data that they make available to administrators and analysts at the University, and answer common questions related to their data.

• What do you mean by that?

November 14, 2019

Presenter: Cynthia Carlton, University Information Architect

Is everyone speaking the same language around the table? It can feel silly to ask what something means; especially when that thing seems so simple and fundamental. “Everyone knows what this is!” you think to yourself. But do they? And is that meaning the same for everyone? In this session, we will discuss when and way you should consider defining terms, how to deal with competing contexts, the differences between a business glossary and a data dictionary and best practices you can take away and start using right away on your teams and projects.

• The Data Checklist: Exploring the needs of your data

December 5, 2019

Presenter: Yennifer Hernandez, Business Architect

When adopting a new solution or requesting an IT implementation, do consider your data needs? We will discuss a list of questions that can help determine whether a solution will meet the needs required to capture, manage, and analyze your data successfully.

• Intro to Tableau

January 9, 2020

Presenter: TBD

Everyone is making pretty pictures with their data – how can I do that?�� Tableau is a very popular and very powerful data visualization tool.�� This short introduction will show you what it is, what it can do, and how you might be able to use it for your analytic work or research.�� This session will help prepare new users for more in-depth Tableau sessions scheduled for Love Data Week in February.

• Data Wrangling: How do I get those data into that report?

March 12, 2020

Presenter: TBD

You learned how do to data visualizations – yippee!�� Now what?�� Unless you are only interested in the most simple reports, there is a lot of data wrangling that is necessary to get the data in the shape you need to do the visualization you want.�� Come learn how to munge, wrangle, combine, define, format, and generally tame unruly data to have the right kind of data for clear and accurate reporting.

• So you want to do a survey?

April 16, 2020

Presenter: TBD

It seems like everywhere you turn, someone wants your thoughts: Just a brief survey please! With all the free tools to create simple questionnaires, it’s easier than ever for anyone to collect data on anything. But not all surveys are created equal.�� This session will cover survey basics around who should be surveyed and how to ask the questions that get the answers you want.

• Now what am I supposed to do?

May 21, 2020

Presenter: TBD

Should I email this spreadsheet? Do I know what’s in that chart? What kind of data are these?�� If you have ever been confused about how to handle data appropriately, what data security means, or what it means to use data responsibly, this is the dialogue for you.�� We’ll talk about basic data do’s and don’ts including how to know what kind of data you are working with, how to handle it, and how to use it appropriately.

And in February, be on the lookout for a slew of data discussions during Love Data Week (Feb 10-14, 2020) where we’ll have presentations on Tableau dashboards and reports, data management, the magic of metadata, and how to not be scared of statistics!

Love Data Week Workshops:

• Can’t help falling in love with Tableau?

February 2020

Presenter: Lauren Di Monte, Director, Research Initiatives

Are you interested in learning more about data visualization? Curious about the best ways to tell stories with data? Looking for simple and fast ways to create data dashboards and create compelling reports? In this beginner-level workshop you’ll get a hands-on introduction to creating interactive visualizations with Tableau Public, a free data analysis and visualization tool.

• Letting Your Data Love You

February 2020

Presenter: Adrienne Canino, Science & Data Outreach Librarian

Having trouble finding data for your projects? Struggling to store or share research data? Don’t let data become a four-letter word! In this hands-on workshop you’ll learn simple tools and techniques for effectively managing personal, business, and research data.

• Mad Ģ����ý Metadata

February 2020

Presenter: Maggie Dull, Head of Metadata Strategies

Even if you’re not aware of it, many of your day-to-day activities rely on metadata, or “data about data”. From your phone to the library to Google and beyond, metadata, or structured information, is working hard behind the scenes. Metadata helps you discover new objects, information and ideas, allows you to draw connections between them, and also preserve and manage that information long into the future. Join us to learn more about the basics of metadata, what good metadata looks like, and how good metadata practice can help you in your personal and professional lives.

• [open refine]

February 2020

Presenter: Sarah Pugachev, Research Initiatives Librarian

• [don’t be scared of statistics]

February 2020

Presenter: TBD

The post Introducing Data Dialogues appeared first on Office of the Provost.