Student and staff data: use of data analytics and machine learning

Jen Persson made this Rhyddid Gwybodaeth request to University of Winchester as part of a batch sent to 60 authorities

This request has been closed to new correspondence from the public body. Contact us if you think it ought be re-opened.

Roedd y cais yn llwyddiannus.

Dear University of Winchester,

I am making a request under the Freedom of Information Act 2000 (FOIA).

Please can you provide information pertaining to any data analytics software used by the university using personal data about your students and staff, since 2012.

1a). Whether you are using any data analytics software at all:
1b) if so, which ones?
1c) what date this data analytics processing began?
1d) indicate whether in the past and ended, or still in current use

2a). Any predictive analytics software such as used in but not only, Blackboard, Xray and JISC supported Civitas.
2b) if so, which ones?
2c) what date this analytics processing began for predictive purposes?

3. If the answer to Q1a or Q2a is yes, please provide a copy of any contract terms on using the data by the provider company including
a) the data items feedling analytics, any profiling, and/or used in prediction
b) what the prediction purposes / aims are
c) training the software (such as AI and machine learning) by the provider using the data provided via your organisation.

4. Whether the University is using its own student or staff population data for machine learning for purposes not included in 1-3 above? (ie but not only, within the university computing department for any project or with another third-party)

5. From which time periods were the student or staff data collected which are processed in any of the above (ie not only, retrospective student data from the last ten years / only current intake which began in autumn term 2018)?

6. How many people/user accounts have access to the whole population student data in any of the above in this calendar year?

Thank you for your consideration.
Sincerely,

Jen Persson

Melanie.Short, University of Winchester

Dear Jen,

Thank you for your FOI request. I have forwarded this onto the relevant department and will reply to you within the 20 day time period.

Kind regards,

Melanie Short
Interim Data Protection Officer
The University of Winchester

dangos adrannau a ddyfynnir

Melanie.Short, University of Winchester

Dear Jen,

In relation to question 1 please could you clarify what you mean by 'data analytics software' and give some examples of this type of software?

For question 4, would machine learning techniques used to determine if an uploaded staff or student photograph will be approved count in what you are looking for?

Kind regards,

Melanie Short
Deputy to the Data Protection Officer
The University of Winchester

dangos adrannau a ddyfynnir

Dear Melanie.Short,

Any predictive analytics software such as used in but not only, Blackboard, Xray and JISC supported Civitas.

Data analytics refers to the measurement, collection, analysis and reporting of data about the progress of learners and the contexts in which learning takes place, or associated data processing of student or staff data.

Regards your question on AI: any AI processing student or staff data so if this is used to process photographs, yes this would fall into this category.

Sincerely,

Jen Persson

Melanie.Short, University of Winchester

Dear Jen,

Thank you for your clarification, please find the response to your FOI request below:

1a). Whether you are using any data analytics software at all:

We are not using any data analytics software

2a). Any predictive analytics software such as used in but not only, Blackboard, Xray and JISC supported Civitas.

We are not using any predictive analytics software

3. If the answer to Q1a or Q2a is yes, please provide a copy of any contract terms on using the data by the provider company.

N/A

4. Whether the University is using its own student or staff population data for machine learning for purposes not included in 1-3 above? (ie but not only, within the university computing department for any project or with another third-party)

We are trialling the use of machine learning for photo upload approval. Currently we are training the system to learn from previous rejected photos and then we will move onto testing the handling of photo upload approvals within the next 3 months

5. From which time periods were the student or staff data collected which are processed in any of the above (ie not only, retrospective student data from the last ten years / only current intake which began in autumn term 2018)?

Only the current intake which began in autumn term 2018

6. How many people/user accounts have access to the whole population student data in any of the above in this calendar year?

Only a small number of staff within IT Services - no more than 15 people/user accounts due to the nature of the work being done with machine learning.

Kind regards,

Melanie Short
Deputy to the Data Protection Officer
University of Winchester

dangos adrannau a ddyfynnir