03 October 2018 Minutes

Meeting minutes of the first Board of Studies meeting for 2018/19 held at 2pm on Wednesday, 3rd October 2018 in Informatics Forum, Room 2.33.

PLEASE NOTE THAT THESE MINUTES HAVE YET TO BE APPROVED BY THE CONVENER AND ARE SUBJECT TO CHANGE.

 

Present - Stuart Anderson, Alan Smaill, Adam Lopez, Frank Keller, Alex Lascerides, Sharon Goldwater, Vicky Mactaggart, Guido Sanguinetti, Pavlos Andreadis, Nigel Topham (late arrival), Alexandra Welsh.

Convenor gave apologies from: Jane Hillston, Steven Renals.

Draft Minutes of Previous Meeting - 18 July 2018: have not yet been distributed as the convener has not approved them. The minutes refer to the New Proposed Programme - BSc Data Science (Graduate-Level Apprenticeship) and related courses; Work-based Professional Practice (A) in Data Analytics; and Work-based Professional Practice (B) in Data Science.

Matters Arising: Stuart Anderson advised that he would read the draft minutes after this meeting and provide approval. Convener approved minutes to be published on the web page and circulated. 

ACTION: Alexandra Welsh to upload minutes from 18th July 2018 on Board of Studies web page.

 

Item 1) New Proposed Programme - CDT in Biomedical AI (PhD) - Guido Sanguinetti

This item is a new CDT programme, which has not yet had reviews but needs approval.  PhD would look to start in 2020 with no direct access. MSc to start in 2019. New courses need to be created which would be project based. It was highlighted that cost tables would be looked at as part of stage 2 approval phase. Guido requested guidance on this which Stuart advised would be organised.

It was discussed that this programme wouldn't need more than standard alternative awards, with the only difference being the hurdles of entry and allocating supervisors. There would also be a hurdle of 65% for project marks. Stuart noted that the phrasing of the programme statement is tricky as if the student doesn't show themselves as suitable, they fail. Adam Lopez feels that having marking criteria helps. Sharon agreed and added that Learning Outcomes is also a useful tool - typical or matrical.

It was agreed that minimal learning outcomes would need to be met in order to obtain award and that the same could be done with marking requirements that feed progression.  It was also discussed that discretion could also be used to progress a student if the progression board feels it is ok to do so; i.e. if a student's quantitative measures were not met, but they did meet other measures and were able to come through.

OUTCOME: Approved pending amendments and addition of statements.

ACTION: Guido to update proposal documents with agreed hurdles and statements (1.5 Programme Structures in the Stage 2 element). Adam Lopez and Stuart Anderson to help with this.

ACTION: Stuart Anderson to help Guido with the cost table in Stage 2 document as Guido is not sure how to complete numbers for Home and International Students.

 

 

Item 2) New Proposed Programme - UKRI CDT in Biomedical AI (MSc Res) - Guido Sanguinetti

It was discussed that the the bid cannot be changed too much as it has been submitted. The programme would have 60 taught credits (40 credits as part of a group project) and 80 credits in an individual project.  It was discussed that a DPT would need to be prepared, which may be broad but will need to include analytical skills. The programme also needs ethical directors and Biomedical involvement.

The programme would also include a course series of lectures from clinicians, for which Guido already has a letter of support. Students would also shadow researcher/mentors on the programme. The programme already has fixed funding.

There was some concern around competitors as there are already 2 similar programmes. Sharon Goldwater also questioned the credits of the course and how the options would work with ofther schools, such as Mathematics. An example of credit breakdown could be 20 credit research, 20 credit bio and 20 credit MLPR. Guido reassured that there is flexibility in the DPT.  Realistically, all students would take Machine Learning courses. There were concerns regarding exit points for students who might not pass all of the required hurdles, however it was discussed that these students would then graduate with an equivalent diploma or certificate.

Adam Lopez feels it is a complicated issue as there is no progression through MSc. PhD students don't know their marks until 1 month after the PhD programme has started. It was agreed that Adam and Guido would have a discussion and that Adam would advise on a state of progression bar.  It was highlighted that PePar has similar issues, however there are clear hurdles that the students have to meet and a progression meeting where the students progress onto a programme is considered. This disclosure needs to be added to the programmes terms and conditions on the proposals. It was also agreed that students cannot progress without having a supervisor and that supervision is in place.  It was discussed that another disclosure needs to be added to the programme proposal whereby it states that the school reserves the right to exclude the student from the programme if the school is unable to provide supervision. This is already in place for normal PhD students and standard regulations cover this.

OUTCOME: Approved pending amendments and addition of statements.

ACTION: Guido to update proposal documents with agreed hurdles and statements (1.5 Programme Structures in the Stage 2 element). Adam Lopez and Stuart Anderson to help with this.

 

Item 3) New Proposed Course - Biomedical Artificial Intelligence MSc project - Guido Sanguinetti

Sharon was concerned that the Learning Outcomes are roughly the same objectives for each course; need more measurable learning outcomes for individual courses. It was discussed that staff might require further guidance on learning outcomes and objectives. Alexandra Welsh advised that she attended some IAD training at which they distributed Learning Outcome guides and that she had uploaded them to the Course Proposal Infweb page along with the updated SQFC course and programme guidance.  Sharon reported that some of these link are broken, so need to be reviewed.

Stuart feels it would be beneficial if all the CDT courses and programmes had similar names, so they can be uniform with the existing CDT. It was discussed that acronyms for all courses in Informatics should also make sense. All future proposed course/ programme names and acronyms for Informatics should be consulted with Vicky Mactaggart, from an admin perspective, and Sharon Goldwater, from a curriculum perspective.

OUTCOME: Approved pending amendments and addition of statements.

 

ACTION: Guido, Adam and Frank to come up with names in consultation with Vicky Mactaggart and Sharon Goldwater.

ACTION: Guido will check statements and Learning Outcomes to put on programmes with Vicky Mactaggart.

ACTION: Alexandra Welsh to send an email via staffmail with IAD guides to Learning Outcomes and  links to the Board of Studies Course Proposal guidance page.

 

 

Item 4) New Proposed Course - Group project in Biomedical AI - Guido Sanguinetti

In this course the student would be paired with a 4th year Biomedical students to write a collaborative report, made up of 25% report & 25% joint presentation. It was discussed that therev needs to be a clear distinction of individual participation. It was discussed that the goal is to have 12 students do a peer assessment. Sharon questioned if all students have to do training in research methods? Adam advised that there is already explicit training in CDT so as to build up god tutors in the subject. It is an expectation in the CDT proposal. The idea is that PhD students are supported in the 1st year on the understanding that they give back as there is a massive Teaching Support resource issue.

Guido is looking at a compressed version of a tutorial. Alex Lascarides advised that there is a list of recommended courses run by the IAD each year that the PhD should do. This would be over and above supervisor support and although not compulsory, is strongly recommended. Alex Lascerides commented that VOX did a writing course which could work and that she thought was very beneficial to the students, but is expensive. The VOX course is comprised of 1/2 day course in the afternoon covering how to write. The following morning of the course is comprised of creating a write up, which is then verbally presented in the afternoon for feedback. Alex feels that the model cannot be mirrored as you get what you pay for with the VOX staff and their experience and expertise. Course costs £10,000.  Stuart agreed that, since the ability to write is critical across the programmes Alex Lascerides should look at obtaining funding for writing courses.

OUTCOME: Approved pending amendments.

ACTION: Alex Lascerides to investigate ways to strengthen writing in graduate school.

ACTION: Guido to get guidance on full course proposal and group disputes.

 

Item 5) New Proposed Course - Issues in Clinical Data Modelling - Guido Sanguinetti

Guido advised that this is a special course that is essential - it is co-ordinated by him but given by Biomedical. Students on this course would be assigned a lecturer to shadow. The course is broken up into 70% course work and 30% presentation. This course wouldn't be part of the Informatics course collection.

Clarification is required around who the host lecturer will be and how it fits in with the assessment - will there be guest lecturers? Guido advised that it would work the same way as the MSc dissertation and only makes up 10 points. It was discussed that the course organiser would need to oversee the guest lecturer.

It was discussed that there might be an imbalance if there is an interesting lecture or topic. It was discussed that Guido would need to look at how course work topics are framed. Students will be assigned to a subject for their 10 page write-up, they might not have complete free choice.

OUTCOME: Approved pending amendments and restrictions.

ACTION: Guido to remove course collection.

ACTION: Guido to restrict students to 3x topics - students may not have complete free choice.

 

Item 6) New Proposed Programme - UKRI CDT in Natural Language Processing - Adam Lopez

This is a new form of PhD with Integrated Study for the CDT that takes place over 4 years. There is already a similar programme on the books in Medicine, and that’s what this programme was modelled on.  Adam and Frank have submitted both Stage 1 & Stage 2 Programme Proposals together. Section 1.5 of the Stage 2 document shows the grad phase in PhD research. Students would take courses in years 2 & 3 to distribute workload. There is an example of the DPT in the documents submitted for consideration containing 40 points of foundational courses, 20 points of group projects (not an MSc project), 40 points of individual project and 60 points of PhD research in the first year. Specialist courses begin in year 2 through to year 3. Year 4 would be made up of PhD research only. The programme is made up of a total of 180 points per year, 150 points of which have to be at level 11 or above.

The entry requirements of the programme have been copied from the existing programme in Medicine. The course results would be examined by MSc board and have annual report and would act as a progression to PhD research as PhD research doesn’t have the same threshold as years 1-3. It was discussed that the progression hurdles need to be looked at as progression may be affected if the students hasn’t passed all of their courses, but could pass on aggregate. Adam reassured the board that the University wide programme structure guide is being used and is compliant with the regulations for Integrated Studies. Alternate awards would be taken from existing programme (MSc R – Informatics) as there is no existing degree on subsidiary of programme. Students would then do a 180 point dissertation and if pass, will graduate in July.

It was clarified that all courses on the programme will be available on EUCLID, except the three new ones being proposed specifically for this programme. 1st year courses are in red on the DPT table and the 2nd year courses are in blue; existing courses are in black.

It is unclear how this would be shared with PPLS – there would need to be a work flow and understanding. It was discussed that there is a precedence with shared programmes/courses relating to CDT and Maths and that the same method would be used with PPLS.

OUTCOME: Approved. To be submitted to College.

ACTION: ISS to submit proposal to College.

 

Item 7) New Proposed Course - Individual Project in Advanced Natural Language Processing (IPANLP) - Adam Lopez & Frank Keller

Items 7 & 8 were considered together as they relate to similar courses; Individual project in advanced NLP and Group project in advanced NLP. Both were submitted as full course proposals to this Board of Studies. The course, Individual Project in Advanced Natural Language Processing, will only be available to students on the 1st year of the PhD in NLP. There would be formative assignments evaluated by peer assessment at group meetings. Course work is modelled on informatics honours project (PGT MSc thesis).

OUTCOME: Approved. To be submitted to College.

ACTION: ISS to submit proposal to College.

 

Item 8) New Proposed Course - Group Project in Advanced Natural Language Processing (GPANLP) - Adam Lopez & Frank Keller

The course, Group Project in Advanced Natural Language Processing, will only be available to students on the 1st year of the PhD in NLP. Look at running the course in semester 1, with meetings occurring about five times over the course of the semester. The course will include formative assignments evaluated by peer assessment at group meetings. Potentially look at conference or workshop papers. This course is also modelled on informatics honours project (PGT MSc thesis), though not to necessarily true to scale. Look at courses being created for start of academic year 2019/20.

OUTCOME: Approved. To be submitted to College.

ACTION: ISS to submit proposal to College.

 

Item 9) New Proposed Course - Doing Research in Natural Language Processing (DRNLP) - Adam Lopez & Frank Keller

This course, Doing Research in Natural Language Processing, is similar to existing Informatics courses, Informatics Research Review (INFR11136) or Informatics Project Proposal (INFR11147). Input has been sought from Björn Franke regarding group and individual projects. Proposed course would run in both semesters 1 and 2, with students meeting once per week. This course would only be available to students on the 1st year of the PhD in NLP. Methods Assessed, includes 2x write-ups (1x draft of PhD proposal and 1x other). It was clarified that one person would be the course organiser for all 3x proposed courses as the courses are not work heavy. Stuart reminded that assessments should be tied back to Learning Outcomes.

OUTCOME: Approved to send Stage 2 to College with Stage 1.

ACTION: Proposal to be submitted to College. Stuart agreed to present the proposals to College if Adam and Frank cannot be present.

 

Item 10) New Proposed Course - Computer Design and Computer Architecture - Nigel Topham

This proposal to merge two existing UG3 courses, Computer Design (INFR09046, acronym CD) and Computer Architecture (INFR09009, acronym CAR), and forms part of the overhaul of Informatics courses primarily to reduce the number of courses offered. These courses contain some overlap of material, and therefore are a natural choice to combine. This would create a 20 point, level 10 course, entitled Computer Architecture and Design. Merger will reduce number of taught courses by 1, with total points taught will reduce from 30 to 20. The assessment structure is based on the structure currently used in the Computer Design course, with coursework worth 40%. There would also be intensive labs and tutorials in the 2nd part. Stuart and Sharon both feel that this is a good proposal.

It was discussed that the proposal forms need to be reviewed to help the proposer distinguish which parts are for admin purposes and which parts will be shown on the DRPS course descriptor. Alexandra Welsh advised that the forms are already in the process of being revamped using a webmark form that combines both the EUCLID information required and the current full course proposal word document.  These changes have been sought to improve the information being input on EUCLID from the proposals being submitted; and have previously been discussed with Stuart Anderson. It was clarified that there is a regulatory body that governs what is promised in what the school advertises.

OUTCOME: Approved. To be submitted to College pending Full Course Proposal.

ACTION: ISS to submit Full Course Proposal  to College.

ACTION: Sharon Goldwater to meet with Alexandra Welsh to look at course proposal forms and highlight which parts will show on DRPS course descriptor.  - COMPLETED

 

Item 11) For Information - Focusing/differentiating the MSc degrees - Sharon Goldwater

 In order to manage student numbers on courses we need clearer ‘rights’ relating to programmes and the courses students can take. This should help improve the experience of staff and students by better managing student class sizes.

Sharon has consulted Stephen Gilmore, Stuart Anderson and Vicky Mactaggart on this and received feedback.  The focus would be to change the Online Degree Finder and focus the requirements of the degrees. The board approved the proposal to manage student numbers and provide a more efficient experience.

OUTCOME: Approved.

ACTION: Changes to be made to degrees by ISS to appear on Online Degree Finder & DRPS for 2019.

 

Item 12) For Information - Proposed Sem 2 instance for Postgraduate Course: Introductory Applied Machine Learning (INFR11182 ) - Nigel Goddard

Item was struck from the agenda as no papers were submitted to explain the proposal.  Stuart clarified that Informatics need to put an S2 version of IAML on the systems for 2019/20, so it can be included in the DPT for the FinTech MSc that is being run out of the Business School/EFI.  This course would basically be 11182 run in S2 and restricted in the first year to the FinTech MSc students. 

It was clarified that the course name needs to be made clear from the start and needs to make sense as a new course name would mean creating a whole new course on EUCLID. This applies to levels and credits of a course too. College of Science and Engineering do not allow caps on courses.

ACTION: Stuart to speak to Nigel and FinTech MSc organiser about creating a new version of 11182 in the business school, so they can own it and it wouldn’t show on an Informatics Course List.

 

Item 13) For Information - Flipping Data Mining and Exploration (INFR11007) - Arno Onken & Bob Fisher

Item was struck from the agenda as no papers were submitted to explain/justify the proposal.  Stuart did advise, however, that flipping a course is a teaching decision - it is not likely to be approved if the board doesn’t feel it is appropriate. There is an agreement not to flip anymore courses. It is unclear as to the proposer’s intention or understanding is as he is not here to present the item.

It was discussed that the Informatics course Performance Modelling (Level 11) was flipped approved to be flipped at a Board of Studies meeting in October 2017 under plans for DSTI (Data Science, Technology and Innovation (DSTI) interdisciplinary programme), however members recalled there was some dissidence at that meeting. It was noted that the Board of Studies members are not happy approving distance learning/flipped courses.

ACTION: Stuart to contact Bob Fisher and Arno Onken with regards to the flipping and the distance learning conversion of courses.

 

Item 14) For Information - Programmes that were updated/changed over summer 2018

  1. Data Science (MSc) & Data Science (MSc)(R) Degree Programmes replaced 3x Maths courses that are no longer running  (Nonlinear Optimization  MATH11031 10 credits, Combinatorial Optimization MATH11030  5 credits, Stochastic Optimization MATH11010 5 credits), with 2x courses; Integ and Cominatorial Optimisation MATH11192 10 credits & Topics in Applied Optimisation MATH11194 10 credits.
  2. Informatics (MInf) (UTINFMT) Degree Programme removed hurdle wording "Entry to Third Year requires passes at 50% or above at the first attempt in at least 40 credit points' worth of the Informatics 2 courses listed together with passes in all remaining compulsory courses, or the permission of Head of School." from the Year 2 2018/19 DPT. (Note: Yr 3 to Yr 4 hurdle is to remain as is).

 

Item 15) For Information - Courses that were updated/changed over summer 2018

  1. Introduction to Vision and Robotics (INFR09019) now has a December exam diet, and an August Resit. The April/May diet option has been removed.
  2. Data Mining and Exploration (INFR11007) removed Probabilistic Modelling and Reasoning (INFR11134) as a c o-requisites.
  3. Extreme Computing (INFR11088) change to assessment weightings from 50% for two assignments; 50% for the exam to 40% for a single assignment; 60% for the exam. There will be no more practical labs.
  4. Introduction to Modern Cryptography (INFR11131) moved course from Semester 1 to Semester 2 for academic session 2018/19.
  5. Computer Programming Skills and Concepts (INFR08022) assessment weightings changed from 90% exam; 10% coursework to 80% exam; 20% coursework.
  6. Algorithmic Game Theory and its Applications (INFR11020) assessment weightings changed from 70% exam; 20% coursework to 80% exam; 20% coursework.
  7. Internet of Things Systems, Security, and the Cloud (IoTSSC) (INFR11146) moved Computer Communications and Networks (INFR10074) from pre-requisites to co-requisites.
  8. Advanced Vision (Level 11) (INFR11031) & Advanced Vision [DL] (INFR11151) changed assessment weightings from 25% coursework; 75% exam to 30% coursework; 70% exam.
  9. Principles and Design of IoT Systems (INFR11150) deleted Computer Communications and Networks (INFR10074) as a pre-requisite and moved IAML INFR10069 to be a co-requisite.
  10. Informatics Distance Learning Courses removed from Informatics DPTs -
  • Advanced Vision INFR11151
  • Introductory Applied Machine Learning
  • INFR11152, Introduction to Java Programming INFR09050
  • Image and Vision Computing INFR11155
  • Introduction to Vision and Robotics INFR11153
  • Natural Computing INFR11165
  • Performance Modelling INFR11154

 

AOCB (Any Other Current Business)