Recommender Systems M

Assessed Coursework

Course Name: Recommender Systems M

Coursework Number: 2

Deadline Time: 4.30pm

Date: 4th August 2021

% Contribution to final course mark: 8%

Solo or Group ✓ Solo: X

Group:

Anticipated Hours: 8-10 hours

Submission Instructions: As per specification below.

Please Note: This Coursework cannot be Re-Assessed

Code of Assessment Rules for Coursework Submission

Deadlines for the submission of coursework which is to be formally assessed will be published in course documentation, and work which is submitted later than the deadline will be subject to penalty as set out below. The primary grade and secondary band awarded for coursework which is submitted after the published deadline will be calculated as follows:

(i) in respect of work submitted not more than five working days after the deadline

a. the work will be assessed in the usual way;

b. the primary grade and secondary band so determined will then be reduced by two secondary bands for each working day (or part of a working day) the work was submitted late.

(ii) work submitted more than five working days after the deadline will be awarded Grade H.

Penalties for late submission of coursework will not be imposed if good cause is established for the late submission. You should submit documents supporting good cause via MyCampus.

Penalty for non-adherence to Submission Instructions is 2 bands

You must complete an “Own Work” form via

https://studentltc.dcs.gla.ac.uk/ for all coursework

Recommender Systems (M)

Exercise 2 – Coversheet

Introduction

The aims of this exercise are:

• Implement a user-based collaborative filtering technique

• Instantiate and evaluate matrix factorisation recommendation models for both explicit and implicit factorisation.

• Develop and evaluate baseline recommender systems

Exercise Specification

Use the provided Colab template notebook (see Moodle). The notebook is structured into Tasks, designed to familiarise yourself with a recommender system dataset, and apply Pandas to solve recommender systems tasks. Once you conduct a task, you should answer the corresponding questions on the Exercise 2 Quiz instance. Ensure that you click the “Next Page” button to incrementally save your answers on the Quiz instance.

When implementing code, you are advised to comment your code solutions, using Python comments and/or markdown cells.

Finally, upload your completed notebook (including all your solutions and the results of the execution of your solutions) and answer the concluding questions of the Quiz. Please note that a 2-bands penalty will be applied, if you do not upload your completed notebook or if you do not show all the results (inc. plots) obtained from the execution of your solutions.

This exercise is graded out of a total of 50 marks, which will be converted to a grade & band.

The marks are allocated to Tasks as follows:

Task 1 (Similar Users) – 4 marks

Task 2 (User-based CF) – 4 marks

Task 3 (Mean-centre User-based CF) – 4 marks

Task 4 (Examining MF latent factors) – 4 marks

Task 5 (MF tuning) – 5 marks

Task 6 (Baseline Recommenders) – 8 marks

Task 7 (Artist Analysis) – 4 marks

Task 8 (Implicit Recommendations) – 5 marks

Task 9 (Recommendations and Listen Analysis) – 8 marks

Task 10 (BPR) – 4 marks

Hand-in Instructions

All your answers to Exercise 2 must be submitted on the Exercise 2 Quiz instance on Moodle with your completed notebook (showing both your solutions and the results of their executions). The submitted notebook will be used to spot check your answers in the Quiz. This exercise is worth 50 marks and 8% of the final course grade.

NB: You can (and should) naturally complete the answers to the quiz over several iterations. However, please ensure that you save your intermediary work on the Quiz instance by clicking the “Next Page” button every time you make any change in a given page of the quiz and you want it to be saved.