COMP3771 User Adaptive Intelligent Systems
Coursework (30% of module)
Deadline for submission: 10am Monday 22 November 2021
Submit an electronic copy in Gradescope following the link in MINERVA
Learning objectives
- to apply the schema of a user-adaptive system in a specific context;
- to design a user-model for adaptation;
- to assess the strengths and weaknesses of your proposal;
- to gain a deeper understanding of the lectures and readings on general architecture of useradaptive
systems and user models; - to practise technical report writing.
Task
Propose a conceptual design of a personalised wellbeing assistant and its underpinning user
model. The vision for this personalised wellbeing assistant is to offer relevant and timely
suggestions / advice helping people to have a healthy lifestyle and feel well.
Your application should focus on one wellbeing aspect and consider a specific user group. For
example, you may consider students or people with special needs (e.g. people with disabilities),
or people from different demographics (e.g. children, elderly, people moving to a foreign
country). Possible wellbeing aspects include (but are not limited to): mental health, activity,
fitness, diet, sleep, stress, social life, acts of kindness, etc.
The proposal should be a maximum of 1500 words (excluding illustrations and references).
You should indicate the word count at the start of your report.
Coursework specification
Your proposal should be written up in an academic style and in a report structure with
appropriate sections (as listed below), and should include appropriate illustrations and
references.
The content of the report should include the following: - An introduction to the wellbeing aspect and the specific user group with clear
justification why you have selected this aspect and this user group (with appropriate
references) and why personalisation is needed. The introduction should also define
clearly the purpose and scope of your personalised wellbeing assistant. - A user scenario that illustrates two user-adaptive features in your proposed
personalised wellbeing assistant. - A description of the user model with clear indication what will be included in the user
model and how it relates to the two user-adaptive features. - A description of how the user model can be built – describe how information can be
captured and represented (with appropriate illustrations); consider both explicit and
implicit user profiling, consider context modelling, and explain how the data collected
about the user will be used to build the user profile. - An overall architectural diagram of the key components for the proposed two useradaptive
features of the personal wellbeing assistant. The graphical representation
should be supported by a brief explanation of each component and its input and output.
Note that you are NOT required to describe in detail the algorithms in each component. - A critical review of the proposal, discussing one strength and one potential limitation
for each of the two user-adaptive feature you have proposed, including both a
computational perspective and a human factors perspective.
Resources
• Re-read articles issued in the module as you may find helpful ideas from these experts
on the overall architecture, the design of the user model, and the user modelling
methods to be used.
• Visit User Modeling Inc (https://www.um.org/) for examples of user-adaptive systems
presented at the past conferences and the journal. Visit the Recommender Systems
conferences (https://recsys.acm.org/). Reference the systems that inspire the
personalisation features to include in your application. Your inspiration may come from
another domain and can be adapted for the problem in this coursework.
• Use online digital libraries, e.g. ACM digital library (https://dl.acm.org/) or IEEE
Xplore (https://ieeexplore.ieee.org/).
• The University of Leeds Library has academic skills links
(https://library.leeds.ac.uk/i…) which can be helpful, e.g.
referencing and searching for information.
• The Institute of Engineering and Technology resource on report writing
(https://www.theiet.org/media/…) provides useful tips.
Marking scheme
Section Points Marks
available
Section 1 –
Introduction
Introduces: - the chosen wellbeing aspect with justification
- the selected user group with justification
- justification why personalisation is needed
- purpose clearly defined
- scope clearly defined
16
Section 2 –
User scenario
A user scenario that illustrates two user-adaptive features: - feature 1 – appropriate and illustrated
- feature 2 – appropriate and illustrated
- suitability of scenario
12
Section 3 –
User model
description
A description of the user model, including: - outline of the user model
- detail what is included
- appropriateness of the model for feature 1
- appropriateness of the model for feature 2
16
Section 4 –
User model building
A description of how the user model can be built, including: - how information is captured
- how information is represented
- how the collected data is used to build the user model
- appropriate illustrations
16
Section 5 –
Overall architecture
An overall architectural diagram of the key components: - main components clearly specified
- the architecture is suitable for the two features
- explanation of main components is appropriate
- clear indication of the input/output of each component
16
Section 6 –
Critical review of the
proposal
A critical review of the proposal: - feature 1 strength
- feature 1 limitation
- feature 2 strength
- feature 2 limitation
- clear indication relevance to computational and
human factors perspectives - both perspectives included